Sigma Plot 11 Users Guide
Sigma Plot 11 Users Guide
Introduction 1 Notebook Manager Basics 29 Worksheet Basics 41 Creating and Modifying Graphs 107 Graph Page Basics 235 Working with 2D Plots 309 Working with 3D and Contour Graphs 373 Working with Pie, Polar, and Ternary Plots 409 Modifying Axes, Tick Marks, and Grids 421 Using the Report Editor 493 Publishing Graphs 503 Automating Routine Tasks 515 Using Transforms 571 Nonlinear Regression 667 Editing Code 749 Advanced Regression Examples 773 Transform Function Reference 799
2 Chapter
1 Introduction
1
Introduction
About SigmaPlot
SigmaPlot 11.0 makes it easier for you to present your findings accurately using precise, publication-quality graphs, data analysis and presentation tools. SigmaPlot 11.0 offers numerous scientific options such as automatic error bars, regression lines, confidence intervals, axis breaks, technical axis scales, nonlinear curve fitting and a data worksheet for powerful data handling. SigmaPlot is a state-of-the-art technical graphing program designed for the Windows platform. It is certified for Windows NT, Windows 2000, Microsoft Office 98, 2000, and Windows XP. SigmaPlot is specifically designed to aid in documenting and publishing research, specializing in the graphical presentation of results. Creating and editing graphs is easy. Just click a Graph toolbar button, pick your data with the Graph Wizard, and you can create a graph in seconds. You can create a formatted worksheet, or use templates or the Graph Style Gallery to apply favorite graphs again and again. SigmaPlot 11.0 also includes a powerful nonlinear curve fitter, a huge scientific data worksheet that accommodates large data sets, summary statistics including SigmaStats entire test collection, a mathematical transform language and much more. OLE2 technology is fully supported. You can annotate graphs with the Microsoft Word Equation Editor, edit your graphs directly inside Word or PowerPoint, or plot your data with an Excel spreadsheet right inside SigmaPlot 11.0.
2 Introduction
Templates. The SigmaPlot template notebook contains a variety of page layouts. Apply these predetermined template attributes to previously saved pages and graphs, or create a user-defined template. Store your templates in a SigmaPlot Notebook Template file (.JNT). You may want to create your own template notebook. For more information, see Using Graph Pages as Templates in Chapter 5.
3 Introduction
Graph Defaults
Preset graph attribute default settings, such as size and position, font, and symbol, line and bar settings.
Axis Scales
Create multiple axes for 2D graphs. SigmaPlot 11.0 , by default, automatically calculates axis ranges and enables each plot to contain separate X and Y axes.
Tick Marks. Use both major and minor axis tick marks and grid lines. Tick intervals,
length, direction, thickness, and color are all adjustable; grid line types are also adjustable. Tick labels can be numeric, time series, or customized, using labels in a worksheet column.
Axis Breaks. You can specify an axis break with a different post-break tick interval.
Automatic Legends
Generate legends automatically, or ungroup legends and individually customize text labels.
Smooth Data
Smooth sharp variations in dependent values within 2D and 3D data sets using SigmaPlot 11.0 smoothing algorithms.
4 Introduction
Microsoft Excel
SigmaPlot 11.0 uses automation communication standards to create and open Excel workbooks within SigmaPlot 11.0. This functionality enables you to run transforms, perform statistical tests, and graph data stored in Excel worksheets.
Statistics
Descriptive statistics are available for all your worksheet columns. The Statistics Worksheet lists basic statistics for all worksheet columns. Display linear regression lines with confidence and prediction intervals, chart error bars for graphs of column means, and run paired and unpaired t-tests between worksheet columns. Use the Histogram feature to compute and plot distributions for data sets.
Regression Wizard
The Regression Wizard steps through curve fitting, plotting, and generating a report.
Transforms
Modify and compute data using SigmaPlot 11.0s comprehensive transform language.
Drawing Tools
Change the font, size, and style of any text, and change the color, line type, thickness, and fill pattern of graphs and drawn objects with drawing tools.
Reports
The SigmaPlot 11.0 Report Editor displays regression results and features complete text editing functionality.
5 Introduction
Microsoft Windows XP 2003 and 2007 operating systems. SigmaPlot retains Windows Classic for the Microsoft Windows 2000 operating system.
Global Curve Fitting. For more information, see Global Curve Fitting in Chapter
14.
Improved Quick Transforms. This includes the option to automatically update transforms. This is turned off by default. To turn it on, from the menus select: Tools>Options>General. For more information, see Performing Quick Transforms in Chapter 13. Formatted worksheets. You can create a formatted worksheet by clicking the
6 Introduction
System Requirements
SigmaPlot 11.0 runs under the following systems: Windows 2000 Windows NT 4.0 SP6 Windows XP
Excel Workbooks:
Excel for Office 2000 and 97 takes full advantage of SigmaPlot 11.0s functionality. Import excel workbooks into SigmaPlot.
Hardware:
Minimum requirements are Pentium 200 or better 64MB of RAM 48MB available Hard Disk space CD-ROM drive SVGA/256 color graphics adapter (800 x 600, High Color recommended)
Serial Numbers
This unique serial number is located on the CD cover. Have this number available when you call for product support, payment, or system upgrade. Copy this number to the registration card and send it in to Systat Software, Inc. Registration entitles you to: Unlimited technical support. System upgrades.
7 Introduction
This means that two or more separate users can share SigmaPlot using his or her own set of SigmaPlot files and settings. When someone uses SigmaPlot for the first time, it creates a User Folder just for that person. In this way, many people can use the same version of SigmaPlot without risking damage to others files.
SigmaPlot, or
C:\Documents and Settings\user\My Documents\SigmaStat\Stat11.0 for
For more information, see Using the Graph Style Gallery in Chapter 4.
GraphWizard.ini. This file stores all Graph Wizard settings. For more information, see Creating Graphs Using the Graph Wizard in Chapter 4. HistogramWizard.ini. This file stores Histogram Wizard settings. For more
For more information, see Regression Equation Libraries and Notebooks in Chapter 14.
Template.jnt. This notebook file is where all the graph page templates are stored.
8 Introduction
Similarly, when SigmaStat starts for a new user, it copies user files to the Stat4 user folder: Stat32.ini Stat32.opt GraphWzd.ini Samples.snb
macros Frequency Plot, Power Spectral Density, Rank and Percentile, and Vector Plot. For more information, see Using SigmaPlots Macros in Chapter 12.
Samples directory. This directory includes sample graphs, data and nonlinear
9 Introduction
SigmaPlot Basics
SigmaPlot runs under the Windows operating system and functions within the standard Windows interface. For information on how Windows works, refer to your Windows documentation.
Figure 1-2 The SigmaPlot Desktop
10 Introduction
11 Introduction
The Customize dialog box appears. 2. Click the Toolbars tab. 3. Click New. 4. In the Toolbar name field, type a new name. 5. Click OK. 6. Click the Commands tab. 7. To add a button to the toolbar: Select a category in the Categories box. Drag the command you want from the Commands box directly to your new toolbar. 8. To add a menu to the toolbar: In the Categories box, select Built-in Menus. Drag the menu you want from the Commands box to directly to your new toolbar. 9. Click Close when finished.
12 Introduction
The Customize dialog box appears. 2. Click the Toolbars tab. 3. Select the custom toolbar you want to rename from the Toolbars box. 4. Click Rename. 5. Type a new name in the Toolbar name box. 6. Click OK.
Deleting Toolbars
You can only delete custom toolbars.
To delete a toolbar:
13 Introduction
3. Select the custom toolbar you want to delete from the Toolbars box. 4. Click Delete.
The Customize dialog box appears. 2. Click the Options tab. 3. Under Other, select Large icons. 4. Click Close.
14 Introduction
3. Under Other, select Show ScreenTips on toolbars. Note: If you want to display the corresponding keyboard shortcut key, select Show shortcut keys in ScreenTips as well. 4. Click Close.
2. The Options dialog box appears. 3. Choose the appropriate tab and make changes.
Worksheet. Worksheet options include settings for numbers, statistics, date and time, worksheet display, default column width, number of decimal places, and use of engineering notation. Page. Page options control graph page properties. General. The General tab controls application settings. Report. Set report options, such as measurement units or to display rulers, on the
Reports tab.
Graph. Graph defaults control attributes that are applied to all new graphs. Macro. Select macro options, such as code colors and which macro library to use on
the Macro tab. 4. Click OK to apply the changes and close the dialog box.
15 Introduction
Undoing Mistakes
On the Standard Toolbar, click the Undo button. If you later decide that you didnt want to perform an undo, click the Redo button. These commands also appear on the Edit menu, or you can click Ctrl+Z to undo, or Ctrl+Y to redo. You can perform multiple instances of Undo or Redo.
2D Cartesian Graph
The following figures show a examples of 2D Cartesian graphs available in SigmaPlot. This example of a grouped error bar chart includes: A post break tick interval set to a new value. A Y axis break at 75% along the axis length.
16 Introduction
Error bars using worksheet column data. Bar fill colors using a pattern from a worksheet column. X axis tick labels using text from a worksheet column. A grouped bar chart with specified bar and group widths. Legend symbols and text labels. Image art cut from a paint program and pasted onto the page using the Windows Clipboard.
Figure 4-1 An example of a 2D Cartesian graph.
This example of a multiple scatter plot includes: Common log scale Y axis with major and minor tick marks. Custom sized symbols using an incrementing earthtone color scheme.
17 Introduction
This example of a box plot includes: Variable box widths expressing another variable dimension. Tick mark direction pointing out. X axis tick labels using a category axis scale. A Tukey box plot with mean value lines. Custom fill colors.
18 Chapter 1
This line and scatter plot with error bars includes: A scatter plot of column averaged data points, with Y error bars computed from the standard deviations of the data. A top X axis with tick marks turned off. Y axis with a linear axis scale. A left Y axis title. A left Y axis with major tick marks. Numeric major tick labels. An X axis with a linear axis scale. A bottom X axis title. A right Y axis with tick marks turned off. An automatically generated legend. A reference line.
19 Introduction
20 Chapter 1
21 Introduction
22 Chapter 1
23 Introduction
24 Chapter 1
SigmaPlot Help
SigmaPlots online help uses new HTML online Help. View the HTML Help using Microsoft Internet Explorer version 4.0 or higher.
25 Introduction
SigmaPlot FAQs
Some of SigmaPlots most frequently asked questions (and answers) are available on the Help menu. The SigmaPlot FAQ includes helpful tips and work-arounds.
To view the SigmaPlot FAQs:
Customer Service
If you have any questions concerning your shipment or account, contact your local office. For more information, see Contacting Systat Software, Inc. below. Please have your serial number ready for identification when calling.
Training Seminars
Systat Software, Inc. provides both public and on-site training seminars for Systat Software, Inc. products. All seminars feature hands-on workshops. Systat seminars will be offered in major U.S. and European cities on a regular basis. For more information, see Contacting Systat Software, Inc. below.
26 Chapter 1
In the U.S.: Telephone: (510)-231-4780 (8:00 A.M. to 5:00 P.M. Pacific Time) Fax: (510) 412-2909 E-mail: [email protected] Mail: 1735 Technology Drive, Suite 430, San Jose, CA 95110 In Europe: Telephone: 49 2104 / 95480 Fax: 49 2104 / 95410 Email: [email protected]
Systat Software, Inc. 1735 Technology Drive, Suite 430 San Jose, CA 95110
Tel: 866.797.8288 Fax: 510.231.4789
http://www.systat.com
Outside the U.S.:
27 Introduction
Erkrath, Germany
Tel: +49.2104.9540 Fax: 49.2104.95410
References
We have found the following references very useful for graph design and layout. M. Brent Charland, Ph.D. 1995. SigmaPlot for Scientists. Wm. C. Brown Communications, Inc., 2460 Kerper Boulevard, Dubuque, Iowa, 52001. Cleveland, William S. 1985. The Elements of Graphing Data. Monterey, Calif.: Wadsworth, Inc. (408) 373-0728. Kosslyn, Stephen M. 1994. Elements of Graph Design. New York: W.H. Freeman and Company. Tufte, Edward R. 1983. The Visual Display of Quantitative Information. Cheshire, Conn.: Graphics Press. Available from Science News Books, 1719 N. St. NW, Washington, D.C. 20036. Scientific Illustration Committee of the Council of Biology Editors. 1988. Illustrating Science: Standards for Publication. Bethesda, Maryland: Council of Biology Editors, Inc.
28 Chapter 1
Chapter
Notebook files contain all of your data and graphs, and are organized within the Notebook Manager. This covers: Notebook Manager organization. Saving your work. Creating notebooks and adding notebook items. For more information, see Working with Sections in the Notebook Manager. Opening notebooks and notebook items. Copying, pasting, and deleting notebook items.
29
30 Chapter 2
Each open notebook appears as the top level, with one or more sections at the second level, and one or more items at the third level. Within each section, you can create one worksheet and an unlimited number of graph pages, reports, equations and macros. The most recently opened notebook file appears at the top of the Notebook Manager. Modified Notebook Names. An asterisk next to an item in the Notebook Manager indicates that the item has been modified since the last time you saved the notebook. Notebook Item Names. The default startup notebook is named Notebook1. It contains one notebook section, Section 1, and one worksheet, Data 1. When you save your notebook file, the name of the file appears at the top of the Notebook Manager window. Notebook files use a (.jnb) extension. The default names given to notebook sections and items are, Section (number), Data (number) or Excel (number) and Report (number). Regression equations are named when they are created. New items are numbered sequentially.
1. From the File menu, click Open. The Open dialog box appears. 2. Select a notebook (.jnb) file from the list, and click Open. The notebook appears in the Notebook Manager.
To close a notebook:
1. Select the notebook to close in the Notebook Manager. 2. Right-click, and from the shortcut menu, click Close Notebook. You can also choose Close Notebook from the File menu.
32 Chapter 2
1. To undock the Notebook Manager, double-click the title bar and drag it to the desired location. 2. To dock the Notebook Manager and move it back to its original position, double-click the title bar again. 3. To view summary information, click Show summary information. To hide it, click Hide summary information.
Figure 3-1 The Notebook Manager Displaying Summary Information
4. To collapse the Notebook Manager, click the arrow button on the top right-hand corner of the Notebook Manager when docked. To view again, click the graph icon. 5. To drag and drop the Notebook manager, click the title bar and drag the Notebook Manager anywhere on the SigmaPlot desktop.
1. Click the Save button. The Save As dialog box appears. 2. Navigate to the directory where you want to save your notebook. 3. Type a name for the notebook in the File Name text box. 4. Click Save to save the notebook file and close the Save As dialog box.
To save changes with the same name and path:
Click the Save button on the Standard toolbar. Your file is saved.
To save to a new name and path:
1. On the File menu, click Save As. The Save As dialog box appears. 2. Navigate to the directory where you want to save your notebook. 3. Type a name for the notebook in the File Name text box. 4. Click Save to save the notebook file and close the Save As dialog box.
34 Chapter 2
1. Select one or more items or sections from the notebook. 2. Click the Print button on the Standard toolbar to print the worksheet using all the default settings.
To set printing options before printing a report, graph page, or worksheet:
1. Open each item. 2. Press Ctrl+P. The Print dialog box appears. 3. Click Properties.
Protecting Notebooks
To ensure security of notebook contents, you can lock notebooks using a password. This is particularly useful if two or more users are using the same version of SigmaPlot. You can also use a password to send confidential data to other SigmaPlot users.
Setting a Password
To set a password: 1. Select the notebook in the Notebook Manager. 2. On the Tools menu, click Password. The Set Password dialog box appears. 3. In the New Password box, type a password. 4. In the Reconfirm box, type the password again. 5. Click OK.
36 Chapter 2
1. Right-click anywhere in the Notebook Manager that you want the new section or item to appear. 2. On the shortcut menu click New, and then select the item to create. The new section or item appears in the Notebook Manager.
1. If the summary information is hidden on the Notebook Manager, click View summary information. 2. Select the notebook item and edit as appropriate.
1. In the Notebook Manager, click the section or item you want to rename. 2. Click it a second time. 3. Type the new name. 4. Press Enter. The new section or item name appears. Note: To change the name of the notebook, use the Save As dialog box. For more information, see Saving Your Work above.
38 Chapter 2
2. Choose the appropriate drive and directory of the notebook file to open. 3. Double-click the desired notebook file. 4. If you want to open another type of file, choose the type of file from the Files of type list. 5. Click Open. The opened notebook appears in the Notebook Manager.
1. Right-click the item in the Notebook Manager that you want to copy, and on the shortcut menu, click Copy. 2. Right-click the section where you want to paste the item, and on the shortcut menu, click Paste. The selected item is pasted to the current notebook and section.
40 Chapter 2
Chapter
Worksheet Basics
Worksheets are the containers for the data you analyze and graph. They are spreadsheet-like in appearance but are limited in function, and are column rather than cell oriented. The following figure provides some worksheet definitions:
Figure 3-1 Example of a SigmaPlot Worksheet
To enter data, you can type in, paste, or import data from other sources. You can also automatically generate and place data in worksheet columns by data transforms and statistical procedures. This chapter covers: Setting worksheet display options. For more information, see Setting Worksheet Display Options below.
41
42 Chapter 3
Moving around the worksheet. For more information, see Moving Around the Worksheet below. Entering data. For more information, see Entering Data into a Worksheet below. Importing files from other applications. For more information, see Importing Files from Other Applications below. Exporting worksheets. For more information, see Exporting Worksheet Data below. Viewing worksheet statistics. For more information, see Descriptive Statistics for Worksheets below. Displaying worksheet data. For more information, see Displaying Worksheet Data below. Formatting worksheets. For more information, see Formatting Worksheets below. Cutting, copying, pasting, moving, and deleting data. For more information, see Cutting, Copying, Pasting, Moving and Deleting Data below. Entering and promoting column and row titles. For more information, see Entering and Promoting Column and Row Titles below. Removing outliers and other data. For more information, see Removing Outliers and Other Data below. Using Excel Workbooks in SigmaPlot. For more information, see Using Excel Workbooks in SigmaPlot. Printing worksheets. For more information, see Printing Worksheets below.
43 Worksheet Basics
2. Select Worksheet from the New drop-down list. 3. Select SigmaPlot from the Type list. 4. Click OK. A new worksheet appears.
If you are saving the notebook for the first time, the Save As dialog box appears prompting you for a file name and path for the notebook file. If you are saving the worksheet to an existing notebook file, the notebook is updated to include the new worksheet or the changes to the existing worksheet. To save worksheets as non-notebook files, you must export them.
44 Chapter 3
45 Worksheet Basics
2. Click the Worksheet tab. For more information, see Displaying Worksheet Data below. Options include: General. Select to turn Worksheet undo on or off, or to set SigmaPlot to display an error message if duplicate column titles appear when running transforms. Turn Worksheet undo off if you are using a large data set and have a small amount of memory. Numeric. Select to control how many decimal places you want to appear in the worksheet, or if you want to use E notation. For more information, see Changing Numbers Display below. Date and Time. Select to set the display for the specified columns. For more information, see Changing Date and Time Display below. Statistics. Use the Show and Hide buttons to move the statistics between the Shown and Not Shown lists. These buttons are available only if a Statistics worksheet is in focus. For more information, see Statistics Options below. Appearance. Set column widths, row heights, color and thickness of the worksheet grid lines, adjust data feedback colors, and select a font style and size. For more information, see Displaying Worksheet Data below.
46 Chapter 3
Figure 3-4 The Options Dialog Box Worksheet Tab Data and Time Options
Freezing Panes
You can freeze panes to keep rows and columns visible as you scroll through the worksheet.
To freeze panes:
1. Select a cell below and to the right of where you want the split to appear. 2. From the menus select:
Window Freeze Panes
47 Worksheet Basics
Going to a Cell
You can move the worksheet cursor to any cell in the worksheet by specifying the column and row number in the Go to Cell dialog box.
48 Chapter 3
To go to a cell:
2. Enter the desired column and row number. To select the block of cells between the current highlight location and the new cell, click Extend Selection to Cell. 3. Click OK to move to the new cell.
49 Worksheet Basics
1. Place the cursor in a cell. 2. Type a number, label, or date and time value. 3. Press Enter to move down one row, or use the arrow keys to move around the worksheet. If you make a mistake entering data, click Undo on the Standard toolbar. For more information, see Undoing Mistakes in Chapter 1.
50 Chapter 3
51 Worksheet Basics
52 Chapter 3
Mocha Worksheets. For more information, see SigmaPlot, SigmaStat, SigmaScan, and Mocha Worksheets below. Axon Text and Binary formats. For more information, see Importing Axon Files below. Paradox (.db) Symphony (.wkl, .wri, .wrk, .wks) SYSTAT (.sys, .syd) Microsoft Access (.mdb) When you import data from another application that is left-justified, SigmaPlot assumes it is text.
To import data:
1. Place the cursor to the worksheet cell where you want the imported data to start. 2. From the menus select:
File Import
The Import File dialog box appears. 3. Select the type of file you want to import from the Files of Type drop-down list. 4. Change the drive and directory as desired, select the file you want to read, then click Import, or double-click the file name. Depending on the type of file, the data is either imported immediately, or another dialog box appears.
53 Worksheet Basics
Once you have copied and pasted the data, you can promote the top row of data the variable names - to become the column titles. For more information, see Using a Worksheet Row for Column Titles below.
54 Chapter 3
2. Click the Select ODBC Data Source tab. The User and System Data Sources list contains all defined the ODBC data sources. 3. To add a data source that is not on the list, click Add. The ODBC Data Source Administrator dialog box appears.
Figure 3-8 Adding a Data Source
Tip: Click Help to learn more about the ODBC Data Source Administrators use. 4. Click the User DSN tab. 5. Select a name from the User Data Sources list. 6. Click Add. The Create New Data Source dialog box appears.
55 Worksheet Basics
7. Select a driver for which you want to set up a data source from the Name list, and click Finish. An ODBC Setup dialog box specific to the driver you selected for the data source appears. 8. Enter a name to identify the new data source.
Figure 3-10 Identifying the Data Source
56 Chapter 3
9. Click Select. The Select Database dialog box appears. 10. Select the database, and click OK. 11. Click OK again to close the ODBC Microsoft Access Setup dialog box. 12. Click OK in the ODBC Data Source Administrator. 13. Click OK in the ODBC Options dialog box. 14. If the data source already appears in the User and System Sources drop down-list, select it. The Import Table dialog box appears.
Figure 3-11 Importing Fields from a Table
15. In the Import Table dialog box, select a table from the Select Table/Query drop-down list. 16. Select fields in the table by moving fields from Unselected fields to Selected fields by double-clicking a selection in the list. You can also click << and >> to move all the selections, or < and > to move them individually.
57 Worksheet Basics
17. Click Import to import the fields into the worksheet. The Field names in the database become column headings in the worksheet. All records in the table are imported.
Figure 3-12 Data that has been Imported into a Worksheet
18. To import using SQL, on the ODBC Options dialog box click the SQL Query tab.
Figure 3-13 Setting ODBC Options
19. Under Recently Used SQL, type the name of the path where the SQL is stored, or select a recently used SQL (SigmaPlot Query) from the drop-down list. 20. Click Open to open an .spq file. 21. Click Import to run the query and import the data.
58 Chapter 3
If the SQL is valid, SigmaPlot imports that data into the worksheet based on the SQL statement. Field names in the database become column headings in the worksheet. Only the records defined by the SQL (rows) are imported. If the SQL is invalid, you are prompted to correct the SQL.
3. In the Define Name dialog box, enter a name for the range of data in the Names in workbook box. 4. Follow the steps above for as many data sets that you would like to create, and then save the Excel file. Now you can import this file as a database. For more information, see Importing ODBC Databases above.
59 Worksheet Basics
The Import File dialog box appears. 2. Select an .xls or .wks file to import, and click OK. The Import Spreadsheet dialog box appears.
Figure 3-14 Import Spreadsheet Dialog Box
3. Select either the entire spreadsheet or a specified range of cells. Specify cells using the standard Lotus 1-2-3 notation (for example, A1:C50 for a range from cell a1 to cell c50). 4. When you have finished specifying the range to import, click Import. The selected data is imported. Note: The dialog box indicates whether or not the worksheet is in overwrite or insert mode, and where the imported data will begin. 5. To import spreadsheet data from non-compatible programs, save the spreadsheet as either an Excel or text file, then import that file.
60 Chapter 3
If you want to use an Excel workbook as an actual Excel workbook within SigmaPlot , you must open the workbook instead of importing it. Importing places the Excel data into a SigmaPlot worksheet, and does not open the workbook as an actual Excel workbook. For more information, see Using Excel Workbooks in SigmaPlot.
Note: A quicker method of importing text is copying the data in your source application, then opening SigmaPlot and pasting the data. 1. To specify a different column separator, select Delimiter to activate the delimiter options; then select the appropriate type. You can select commas, hyphens, or any other characters. For example, many databases use semicolons (;) as delimiters.
61 Worksheet Basics
2. To specify a model of the data, use dashes (-) to specify column widths, and bracket characters [ and ] to define the column edges. Use a vertical bar | character to indicate a single-character width column. Click Analyze to re-display the appearance of the file using the new model. 3. To save text import formats, enter a name into the Format scheme box, then click Add. Delete unwanted import formats using the Remove button. 4. To specify a different range, enter the rows and columns to read, then click Analyze. You can use this feature to eliminate file headers and other undesired text. 5. When you are finished specifying the file parameters, click Import. The specified data from the file is imported.
SPSS (.SAV)
If you are importing SPSS (.sav) files, the Import Worksheet dialog box appears prompting you to select variables to import.
To select variables to import:
1. In the Unselected Variables list, select a variable you want to import. 2. Click the single > arrow to move that variable to the Selected Variables list. 3. Click the double >> arrow to move the entire contents of the Unselected Variables list to the Selected Variables list.
62 Chapter 3
4. Click Import to place the data in the SigmaPlot worksheet. Note: SPSS data files use category data as the default data format. For more information, see Plotting Category and Grouped Data in Chapter 6.
63 Worksheet Basics
SAS Data Set (V6) (*.sd2) Minitab (V11) (*.mtw) Minitab (V11) (*.mpj)
The Export File dialog box appears. 3. Select a file format from the Files of type drop-down list, and then enter the file name, directory, and drive for the exported file. 4. Click Export to create the file.
Exporting to SYSTAT
When exporting data to SYSTAT, make sure that there are no text cells or indefinites in data columns you export, or they will be converted by SYSTAT into text instead of numbers.
64 Chapter 3
The running calculations performed for each column appear in a Column Statistics window for that worksheet.
Figure 3-16 Column Statistics Worksheet
Available Statistics
To determine the statistics shown in the Statistics windows, use the Statistics Options dialog box. Most calculations ignore empty cells, missing values, and text. The following statistics appear in the Column Statistics window. Mean. The arithmetic mean, or average, of all the cells in the column, excluding the missing values. This is defined by
65 Worksheet Basics
Std Dev. The sample standard deviation is defined as the square root of the mean of the square of the differences from their mean of the data samples xi in the column. Missing values are ignored.
Std Err. The standard error is the standard deviation of the mean. It is the sample standard deviation divided by the square root of the number of samples. For sample standard deviations
95% Conf. The value for a 95% confidence interval. The end points of the interval are given by:
where
is the mean, s is the sample standard deviation, and t(v,z) is the t statistic for v = n1 degrees of freedom and z = 1.96 standard normal percentile equivalent. 99% Conf. The value for a 99% confidence interval. The end points for this interval are computed from the equation for the 95% confidence interval using z = 2.576. Size. The number of occupied cells in the column, whether they are occupied by data, text, or missing values. Sum. The arithmetic sum of the data values in the column. Min. The value of the numerically smallest data value in the column, ignoring missing values. Max. The value of the numerically largest data value in the column.
66 Chapter 3
Min Pos. The smallest positive value. Missing. The number of cells in the column occupied by missing values, denoted with a double dash symbol (--). Other. Either text or an empty cell.
Statistics Options
To display only a portion of the available statistics, use the Worksheet Options dialog box, then select the column statistics to show or hide. For more information, see Displaying Worksheet Data below.
To specify which statistics are shown or hidden:
67 Worksheet Basics
3. Click the Worksheet tab. 4. Select the statistic(s) you want shown or hidden. 5. Click Show and Hide to move the statistics between the Shown and Not Shown lists. 6. Select the appropriate options to change the column widths and data display.
68 Chapter 3
For example, 1.23 e+03 is 1230, or, equivalently, 1.23 e+03. Select E Notation When Needed or E Notation Always on the Worksheet tab of the Options dialog box if you want to use Scientific Notation. For more information, see Changing Numbers Display below. Engineering Notation, which you can select as an option on the Worksheet tab of the Options dialog box, uses integral powers of 3 (with 10 as the base).
Figure 3-18
69 Worksheet Basics
Figure 3-19 Numbers are displayed in Column 1, dates are displayed in Column 2, and text is shown in Column 3
You can enter numbers, labels, and dates and times directly into the worksheet. You can also convert numbers to dates and times and vice versa. You can change column widths, number decimal places, or date and time format, and you can also change the color and thickness of the worksheet gridlines, and adjust data feedback colors. Note: You can format columns to override the defaults set using the Options dialog box. For more information, see Formatting Worksheets below.
Drag the boundary on the right side of the column heading until the column is the size you want.
To change a row height:
Drag the boundary below the row heading until the row is the size you want.
70 Chapter 3
To adjust column width and row height using the Options dialog box:
The Options dialog box appears. 2. Click the Worksheet tab. 3. In the Settings For list, click Appearance. 4. Set column width and row height in the Column Width and Row Height drop-down lists. 5. Click OK to apply the changes and close the dialog box. SigmaPlot s worksheet can display up to fourteen digits of precision regardless of how many decimal places you specify.
The Options dialog box appears. 2. Click the Worksheet tab. 3. In the Settings For list, click Appearance. 4. Set color and thickness in the Color and Thickness drop-down lists.
71 Worksheet Basics
The Options dialog box appears. 2. Click the Worksheet tab. 3. In the Settings For list, click Appearance. 4. Set data feedback colors and thickness in the X and Y drop-down lists. 5. Click OK to apply the changes and close the dialog box.
72 Chapter 3
2. Click the Worksheet tab. 3. In the Settings For list, click Numeric. 4. Select the number of decimal places from the Decimal Places drop-down list. 5. Click OK to accept the changes and close the dialog box. If the number of decimal places exceeds the column width they appear as # symbols.
73 Worksheet Basics
E Notation When Needed Displays worksheet data as scientific notation only when the length of the value exceeds the width of the cell. The default column width is twelve. 12.00 E Notation Always Always displays data as scientific notation. The number of decimal places is set in the Decimal Places edit box. 12.00e+1 Fixed Decimal Displays data with a fixed number of decimal places. Set the number of decimal places in the Decimal Places edit box. The number of decimal places allowed is limited by the column widththe maximum number of decimal places cannot exceed the column width or it appears as a series of # symbols. The default setting for decimal places is two. 12.00 General Displays data exactly as you enter it in the worksheet. 12
To set the numeric display for your worksheet:
74 Chapter 3
3. Click the Worksheet tab. 4. In the Settings For list, click Numeric. 5. Select a Numeric format setting from the Display As drop-down list. 6. To use engineering scientific notation for worksheet values, select Engineering Notation. For more information, see Engineering and E Notation above. 7. Click OK to accept the settings and close the dialog box.
75 Worksheet Basics
76 Chapter 3
2. Click the Worksheet tab. 3. Select the Date and Time from the Settings for list. 4. Type one of the following examples into the Date box, or select a format from the dropdown list: M/d/yyyy. 10/8/2005 M/d/yy. 10/8/05 MM/dd/yy. 10/08/05 MM/dd/yyyy. 10/08/2005 yy/MM/dd. 05/10/08 yyyy-MM-dd. 2005-10-08 MMMM. Complete month dd-MMM-yy. 08-Oct-05 dddd, MMMM dd, yyyy. Tuesday, October 08, 2005 MMMM dd, yyyy. October 08, 2005 dddd, dd MMMM, yyyy. Tuesday, 08 October, 2005 dd-MMMM-yy. 08-October-05 dd MMMM, yyyy. 08 October, 2005 gg. Era (AD or BC) 5. To change the display Time format, type one of the following examples into the Time box, or select a format from the drop-down list: hh or h. 12 hour clock HH or H. Military hours mm or m. Minutes ss or s. Seconds uu or u. Milliseconds H: h: m: s: or u. No leading zeroes for single digits HH: hh: mm: ss: uu. Leading zero for single digits tt. Double letter AM or PM t. Single letter AM or PM
77 Worksheet Basics
The Options dialog box appears. 2. Click the Worksheet tab. 3. Select a date from the Day Zero drop-down list, or type your own start date. The default start date is 1/1/1900.
78 Chapter 3
Day Zero becomes the number 01.00 when you change from Date and Time to Numbers format. The basic unit of conversion is the day; that is, whole integers correspond to days. Fractions of numbers convert to times. Zero becomes Day Zero, and negative numbers entered into the worksheet convert to days previous to the Day Zero start date. Conversion between date/time values and numbers can occur for the calendar range of 4713 BC to beyond the year 4,000 AD. The internal calendar calculates dates using the Julian calendar until September, 1752. After that, dates are calculated using the Gregorian calendar. Note: If you convert numbers to dates, a start date is applied. If you convert the dates back to numbers, be sure you use the same start date as when you converted them, or they will have a different value.
79 Worksheet Basics
Regional Settings
Drop-down lists in the Options dialog box worksheet tab use the current date/time settings in your operating system. The Windows Regional Settings control date/time delimiters, 12 or 24 hour clock, and AM/PM display. Date and time display formats may be affected by your operating systems Regional Settings. For example, if your Time Zones are specified as British (English), your date values appear as dd/mm/yy. If the setting is US (English), your date values appear as mm/dd/yy. If you want to view or modify the current settings, or view alternative settings available on your system, click the Regional Settings button, or modify them directly from the Windows Control tab. Note: Date and time values appear on the worksheet using the date and time delimiters, generally a forward slash (/) or colon (:). For more information, see Entering Dates and Times above.
80 Chapter 3
When copying worksheet values, values are copied as numeric strings, not date/time. SigmaPlot recognizes Date and Time formats imported from Excel, but you will need to convert most other non-text dates and times from numbers to dates and time.
Formatting Worksheets
You can format entire columns even if they contain no data. If a populated cell in a column is already specifically formatted, as you enter data the entire column continues to use the same format, provided the data is appropriate to that format. When importing data, the import format takes precedence over the column format. Note: Formatting worksheets is not the same as setting worksheet display options. Setting worksheet display options sets the default for the entire worksheet. For more information, see Setting Worksheet Display Options above. You can override these defaults by formatting worksheet columns using the Format Cells dialog box.
To format worksheet columns:
The Format Cells dialog box appears. 3. Click the Data tab. 4. Select a Type. The Type you select determines which Settings are available. Available Types are: Numeric. Select Numeric to control how many decimal places you want to appear or if you want to use E notation in a selected worksheet column. Text. Select text to wrap text using the existing column width.
81 Worksheet Basics
Date and Time. Select Date and Time to set the display for the specified columns. For more information, see Switching Between Date and Time and Numeric Display below.
3. Click the Rows and Columns tab. The selected box reflects the selected block of rows and columns. 4. Set column width and row height from the Column width and Row height drop-down lists.
82 Chapter 3
5. To apply the row and column formats to the whole worksheet, select Apply to entire data region. 6. Click OK to apply the changes and close the dialog box. The worksheet appears with new column and row sizes for the selected cells. Note: Setting row height and column width from the Format Cells dialog box only changes the selected block of data.
1. View the worksheet. 2. Select the data you wish to display in date/time format. 3. From the menus select:
Format Cells
83 Worksheet Basics
4. Click the Data tab. 5. In the Type list, click Date and Time. 6. Select date and time formats from the Date and Time drop-down lists. The sample box changes according to your choice. 7. Click OK. The data is displayed showing the date, time, or date and time as specified. The dates and times that are entered as dates and times are automatically displayed.
84 Chapter 3
Drag the mouse over the desired worksheet cells while pressing and holding down the left mouse button. Hold down the Shift key and press the arrow, PgUp, PgDn, Home, or End keys. Use the Go To command on the Edit menu.
Figure 3-26 Selecting a Block of Data in the Worksheet
To select an entire column, move the pointer to the column title row and click. To select entire rows, move the pointer to the row title column and click.
Pasting Data
To paste data: 1. Click or move the worksheet cursor to the cell where you want to paste the data, or to the upper-left corner of the block.
85 Worksheet Basics
or Click the Paste button on the Standard toolbar, or Press Ctrl+V. Any data in the Clipboard is placed in the worksheet.
Moving Data
Move a block of data by cutting it, selecting the upper-left cell of the new location, then pasting the block. For more information, see Deleting Data below.
Deleting Data
Use the Edit menu Clear command to permanently erase selected data. This operation does not copy data to the Clipboard, and is faster than cutting.
1. Drag the mouse over the region where you want the empty block of cells, column, or row to appear. The selected region of cells indicates exactly which cells will be inserted. 2. Right-click, and then on the shortcut menu, click Insert Cells. The Insert Cells dialog box appears.
86 Chapter 3
3. Select the direction you want the existing data to shift when the cells are inserted, or to insert an entire column or row, select Insert Columns or Insert Rows. 4. Click OK. The column, row, or block of cells appears on the worksheet.
Figure 3-28 The Result of Inserting an Empty Block with Cells Shifted Down
87 Worksheet Basics
1. Drag the mouse over the block of cells, column, or row you with to delete. 2. Right-click, and on the shortcut menu, click Delete Columns. The Delete Cells dialog box appears. 3. Select the direction you want the existing data to shift when the cells are deleted. 4. Click OK.
1. Select the block of data to transpose. 2. Cut or copy the selected data. 3. Select the cell where you want to begin pasting the data, 4. From the menus select:
Edit Transpose Paste
The data is pasted to the worksheet with the column and row coordinates reversed.
88 Chapter 3
1. Double-click the title, and enter or edit the title. 2. Press Enter to accept the new title. The new column or row title appears along with the original column or row number. You must use at least one text character in every column title. If you need to use a number as column title, type a space character (by pressing the space bar) before the number.
The Column and Row Titles dialog box appears. 2. Click the Column tab. 3. Enter the column title in the Title box. 4. To edit an existing title, move to that column by clicking Next or Prev, then edit the title.
89 Worksheet Basics
5. Click OK to close the Column Titles dialog box when you are finished editing column titles.
To enter or edit a row title:
The Column and Row Titles dialog box appears. 2. Click the Row tab. 3. Enter the row title in the Title box. 4. To edit an existing title, move to that row by clicking Next or Prev, then edit the title. 5. Click OK to close the Column and Row Titles dialog box when you are finished editing row titles.
1. If necessary, enter the column titles you want to use in a single worksheet row. 2. Select the cells in the row you want to use as column titles.
90 Chapter 3
4. Click the Column tab. The number of the row you wish to promote appears in the Promote row to titles box. 5. To delete the original row once it has been promoted, select Delete Promoted Row. 6. Click Promote. The selected row contents appear as column titles and the Column and Row Titles dialog box closes.
91 Worksheet Basics
1. If necessary, enter the row titles you want to use in a single worksheet column. 2. Select the cells in the row you want to use as row titles. 3. From the menus select:
Format Column and Row Titles
4. Click the Row tab. The column you wish to promote appears in the Promote column to titles box. 5. Select Delete Promoted Column to delete the original column once it has been promoted. 6. Click Promote. The selected column contents appear as row titles and the Column and Row Titles dialog box closes.
92 Chapter 3
1. Click the cell on the worksheet that you want to promote to a column or row title. 2. From the menus select:
Format Column and Row Titles
The Column and Row Titles dialog box appears. 3. Click the Row tab to promote a row cell to title; click the Column tab to promote a column cell to a title. 4. Click Promote. The content of the cell appears as the column title. 5. Select Delete Promoted Column or Delete Promoted Row to delete the original cell once it has been promoted. 6. Click Next or Prev to move to the next desired column or row, then follow steps above.
1. Find the outlier on the graph, then click it to select the curve, pause, and then click again (do not double-click).
93 Worksheet Basics
2. View the worksheet. The data for the selected symbol is indicated with colored highlighting. Note: It is possible to highlight data points only if you create graphs using symbols.
Figure 3-31 When you find the outlier on the graph, click it once to select it, and click it again, but make sure not to double-click.
94 Chapter 3
5. Click the Data tab. 6. Select Text from the Type list, then click OK. This converts the number to text characters; you can tell this if the alignment of the cell changes to be left aligned.
Figure 3-33 Graph with Removed Outlier
The data point is no longer plotted, and if you perform additional statistics on the graph, the data point will also be ignored.
95 Worksheet Basics
Highlighting Outliers
Another way to remove an outlier is to cut the data and move it to another part of the worksheet. This is useful if you still want to plot the data but ignore the outlier. Then you can plot the moved outlier data with a second plot to continue displaying the outlying data.
To plot outlier data as a separate plot:
1. Identify the worksheet cell(s) corresponding to the outlier(s). 2. Select (highlight) the cells, and press Ctrl+X to cut them. 3. Move to another location in the worksheet and paste the data.
Figure 3-34 Moving Outlier Data to a Different Part of the Worksheet
4. Plot the outlier data by adding it as a second plot to your graph. Change the symbol color or other attributes to distinguish the data.
96 Chapter 3
97 Worksheet Basics
Excel workbooks created by SigmaPlot are initially limited to a single worksheet. Excel workbooks with multiple worksheets that are opened by SigmaPlot as notebooks retain all sheets, but only the first sheet can be used for graphs and statistics.
To open a new Excel worksheet:
The New dialog box appears. 2. From the New drop-down list, select Worksheet.
98 Chapter 3
3. Under Type, click Excel. 4. Click OK. An Excel worksheet appears and is added to the notebook. The Excel Standard Toolbar appears.
1. Close all open Excel workbooks. 2. From the SigmaPlot menus select:
Tools Options
99 Worksheet Basics
4. Select New Notebooks use Excel Workbook. 5. Click OK to apply the changes and close the dialog box. All new notebooks will use Excel workbooks as the default worksheet.
100 Chapter 3
paste the data into a SigmaPlot worksheet. To display the statistics worksheet for the active SigmaPlot worksheet, on the View menu, click Statistics.
To specify page setup functions for the active Excel workbook, on the File menu, click Page Setup to open the Page Setup dialog box. You can modify page, margins, headers and footers, and sheet settings.
Figure 3-37 Setting Printing Options Using the Excel Page Setup Dialog Box
You can export in-place active Excel workbooks to Excels native *.xls file format, as well as any other format supported by Excel.
To export Excel Workbooks:
Excels Save As dialog box appears. 3. Select the desired format from the Save as type drop-down list. 4. Specify the drive and directory in which to save the file. 5. Enter a file name. 6. Click Save to save the file.
Excel Toolbars
An Excel workbook in SigmaPlot always uses Excel toolbar default settings of your last Excel session. You can view any of Excels toolbars by clicking Toolbars on the View menu. Select a toolbar to use from the Excel Toolbars dialog box; the toolbars appear near the workbook window. Note: Switching from or closing an Excel workbook hides any Excel toolbars you may have displayed.
102 Chapter 3
To perform user-defined transforms on an Excel worksheet, use the corresponding column number in place of the column letter that appears in the gray heading area at the top of the column. For example, the transform function: col(1)=data(1,100) corresponds to inserting data values from 1 to 100 into column A of an Excel workbook.
Printing Worksheets
You can print active worksheets by clicking the Print button on the Standard toolbar. You can print any worksheet in a notebook. This section explains:
104 Chapter 3
Printing the current worksheet. For more information, see Printing the Current Worksheet. Previewing worksheets before printing. For more information, see Previewing Worksheets below. Printing column statistics. For more information, see Printing Column Statistics. Setting printing options. For more information, see Setting Printing Options. Configuring printer settings. For more information, see Configuring Printer Settings.
Previewing Worksheets
With a worksheet in view, from the menus select:
File Print Preview
The Print dialog box appears. 3. From the Name drop-down list, select the printer you wish to use. 4. Click OK. The Print Data Worksheet dialog box appears.
Figure 3-40 The Print Data Worksheet Dialog Box for Columns Statistics
5. To print the names of the statistics that appear in the row region of the worksheet, under Headers select Row Headings. 6. Click OK to print.
106 Chapter 3
2. Specify whether you want to print the entire worksheet, only the selected cells in the worksheet, or a specified range of columns by selecting one of the options under Area to Print. 3. Click OK to print the worksheet.
The Print Data Worksheet dialog box appears. 2. Click Setup. The Print dialog box appears. 3. Click OK when you are satisfied with the Printer settings, or click Properties to edit the printer properties. Note: The Properties dialog box options vary from printer to printer.
Chapter
A graph is a representation of selected worksheet columns on a graph page. You select the representation, or graph type (for example, 3D scatter plot, vertical bar chart, and so on), when you create a plot or graph, but you can change it at any time. Most plot types can graph many worksheet columns, column pairs, or column triplets. Depending on the plot type, a separate curve or set of bars represents each column. A graph must have at least one plot, but most graphs can hold many more plots, each with a different type and style. This chapter provides an overview of the graph creation process using the Graph Wizard, including descriptions of the different graph types and styles available, and common modifications. This chapter covers: Setting graph defaults. For more information, see Setting Graph Defaults below. Arranging data for graphs. For more information, see Arranging Data for Graphs below. Creating graphs. For more information, see Creating Graphs below. Creating graphs using templates, layouts, and the Graph Style Gallery. For more information, see Using the Graph Style Gallery below. Modifying graphs. For more information, see Modifying Graphs below. Creating and modifying embedded SigmaPlot graphs. For more information, see Creating Embedded Graphs below. Changing symbol type and other symbol options. For more information, see Changing Symbol Type and Other Symbol Options below.
107
108 Chapter 4
Changing line type and other line options. For more information, see Changing Line Type and Other Line Options below. Changing bar and box widths and spacing. For more information, see Changing Bar and Box Widths and Spacing below. Adding and modifying drop lines. For more information, see Adding and Modifying Drop Lines below. Plotting and solving equations. For more information, see Plotting and Solving Equations below. Plotting and modifying regression lines. For more information, see Plotting and Modifying Regression Lines below. Creating histograms. For more information, see Creating Histograms below.
1. Select the graph. 2. Change its properties using the Graph Wizard, Graph Properties, or other dialog boxes and commands. The graph default options are intentionally limited and simple. If you want to use more complex graph defaults, use templates or the Graph Style Gallery to create complex graphs that can be applied to data as a template, bypassing graph creation entirely. For more information, see Using the Graph Style Gallery below.
To change graph defaults:
2. Click the Graph tab. 3. Change the graph defaults options as desired.
110 Chapter 4
Plots data as XY points using symbols. For more information, see Arranging Data for 2D Plots below.
Line Plot
Plots data as XY points connected with lines. For more information, see Arranging Data for 2D Plots below.
Line and Scatter Plot
Plots data as XY points using symbols connected with lines. For more information, see Arranging Data for 2D Plots below.
Area Plot
Plots data as XY points with regions below or between curves filled with a color or pattern.
Polar Plot
Plots data using angles and distance from center. For more information, see Arranging Data for Polar Plots below.
Ternary Plot
Plots data on a coordinate system based on three different components which always add up to 100%. For more information, see Arranging Data for a Ternary Graph below.
Vertical Bar Chart
Plots data as Y points with vertical bars. For more information, see Arranging Data for 2D Plots below.
Horizontal Bar Chart
Plots data as X points with horizontal bars. For more information, see Creating 2D Plots in Chapter 6.
Box Plot
Plots data as the median and percentiles. For more information, see Creating Box Plots in Chapter 6.
Pie Chart
Plots data as a percent of the total. For more information, see Arranging Data for a Pie Chart below.
Contour Plot
112 Chapter 4
Plots data as XYZ values in 2D space. Format data columns as: many Z; single XY, many Z; or XYZ triplet. For more information, see Arranging Data for 3D Graphs below.
3D Scatter Plot
Plots data as XYZ data points in 3D space. Format data columns as: many Z; single XY, many Z; or XYZ triplet. For more information, see Arranging Data for 3D Graphs below.
3D Line Plot
Plots data as XYZ data points connected with lines. Format data columns as: many Z; single XY, many Z; or XYZ triplet. For more information, see Arranging Data for 3D Graphs below.
3D Mesh Plot
Plots data as a 3D surface. Format data columns as: many Z; single XY, many Z; or XYZ triplet. For more information, see Arranging Data for 3D Graphs below.
3D Bar Chart
Plots data as Z values on an XY grid. Format data columns as: many Z; or single XY, many Z. For more information, see Arranging Data for 3D Graphs below.
Scatter Plots
Simple Scatter
Plots a single set of XY pairs. Format data columns as: XY Pair Single X Single Y
Multiple Scatter
Plots multiple sets of XY pairs. Format data columns as: XY Pairs Single Y, Many X Single X, Many Y Many X Many Y XY Category X Category Y Category
114 Chapter 4
Simple Regression
Plots a single set of XY pairs with a regression line. Format data columns as: XY Pair Single X Single Y
Multiple Regressions
Plots multiple sets of XY pairs with regression lines. Format data columns as: XY Pairs Single Y, Many X Single X, Many Y Many X Many Y XY Category X Category Y Category
Simple Error Bars
Plots a single set of XY pairs with error bars. If using worksheet columns or asymmetric error bar columns, format data columns as: XY Pair; or Single Y If using columns means, first column entry, or last column entry as symbol values, format data columns as:
Single X, Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: Single X, Single Y Replicate Y Replicate
Multiple Error Bars
Plots multiple sets of XY pairs with error bars. If using worksheet columns, asymmetric error bar columns, columns means, first column entry, or last column entry as symbol values, format data columns as: X Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: Single X, Many Y Replicates Many Y Replicates
Simple Error Bars & Regression
Plots a single set of XY pairs with error bars and a regression line. If using worksheet columns or asymmetric error bar columns, format data columns as: XY Pair Single Y If using columns means, first column entry, or last column entry as symbol values, format data columns as:
116 Chapter 4
Single X Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: Single X, Single Y replicate Y replicate If using by category, mean, or by category, median, format data columns as: Category, Many Y
Multiple Error Bars & Regressions
Plots multiple sets of XY pairs with error bars and regression lines. If using worksheet columns, asymmetric error bar columns, columns means, the first column entry, or the last column entry as symbol values, format data columns as: Single X Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: Single X, Many Y Replicates Many Y Replicates If using by category, mean, or by category, median, format data columns as: Category, Many Y
Simple Horizontal Error Bars
Plots XY pairs with horizontal error bars. If using worksheet columns or asymmetric error bar columns as the as symbol values, format as:
XY pairs Single X, Single Y, Many X Many X If using column means, column median, the first column entry, or the last column entry as symbol values, format data as: Single Y, Many X Many X If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: Single X Replicates Single Y, single X Replicates Many X Replicates Single Y, Many X Replicates If using by category, mean, or by category, median, format data columns as: Category, Many Y
Bi-directional Error Bars
Plots XY pairs with both horizontal and vertical error bars. Format data columns as XY pairs. If using worksheet columns or asymmetric error bar columns as the as symbol values, format as: XY pairs Single X Single Y, Many X Many X If using column means, column median, the first column entry, or the last column entry as symbol values, format data as: Single Y, Many X
118 Chapter 4
Many X
Vertical Point Plot
Plots columns of data as Y values. Format data columns as: Many Y Single X, Many Y Many Y Replicates Single X, Many Y Replicates
Horizontal Point Plot
Plots columns of data as X values. Format data columns as: Many X Single Y, Many X Many X Replicates Single Y, Many X Replicates
Vertical Dot Plot
Plots a column of data as Y values. Format data columns as: Many Y Single X Many Y XY pairs X Category
Plots a column of data as X values. Format data columns as: Many X Single Y, Many X YX pairs
Line Plots
Simple Straight Line
Plots a single set of XY pairs connecting the data points with straight lines. Format data columns as: XY Pairs Single X Single Y
Multiple Straight Lines
Plots multiple sets of XY pairs connecting the data points with straight lines. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Many X Single Y
120 Chapter 4
Plots a single set of XY pairs connecting the data points with a spline curve. Format data columns as: XY Pairs Single X Single Y
Multiple Spline Curves
Plots multiple sets of XY pairs connecting the data points with spline curves. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Single Y, Many X
Simple Vertical Step Plot
Plots a single set of XY pairs connecting the data points with vertical and horizontal lines, starting with vertical. Format data columns as: XY Pairs Single X Single Y
Plots multiple sets of XY pairs connecting the data points with vertical and horizontal lines, starting with vertical. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Single Y, Many X
Simple Horizontal Step Plot
Plots a single set of XY pairs connecting the data points with vertical and horizontal lines, starting with horizontal. Format data columns as: XY Pairs Single X Single Y
Multiple Horizontal Step Plot
Plots multiple sets of XY pairs connecting the data points with vertical and horizontal lines, starting with horizontal. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Single Y, Many X
122 Chapter 4
Plots a single set of XY pairs connecting symbols with straight lines. Format data columns as: XY Pairs Single X Single Y
Multiple Straight Lines
Plots multiple sets of XY pairs connecting symbols with straight lines. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Single Y, Many X
Simple Spline Curve
Plots a single set of XY pairs connecting symbols with a spline curve. Format data columns as: XY Pairs Single X Single Y
Plots multiple sets of XY pairs connecting symbols with spline curves. Format data columns as: XY Pairs Many X Many Y Single X, Many Y Single Y, Many X
Simple Error Bars
Plots a single set of XY pairs as symbols with error bars connected with straight lines. If using worksheet columns or asymmetric error bar columns, format data columns as: XY Pair Single Y If using columns means, first column entry, or last column entry as symbol values, format data columns as: X Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: X, Y Replicate Y Replicate
Multiple Error Bars
124 Chapter 4
Plots multiple sets of XY pairs as symbols with error bars connected with straight lines. If using worksheet columns, asymmetric error bar columns, columns means, first column entry, or last column entry as symbol values, format data columns as: X Many Y Many Y If using row means, row median, first row entry, or last row entry as symbol values, format data columns as: X, Many Y Replicates Many Y Replicates
Simple Vertical Step Plot
Plots a single set of XY pairs connecting symbols with vertical and horizontal lines, starting with vertical. Format data columns as: XY Pairs Single X Single Y
Multiple Vertical Step Plot
Plots a multiple sets of XY pairs connecting symbols with vertical and horizontal lines, starting with vertical. Format data columns as: XY Pairs Many X Many Y Single Y, Many X Single X, Many Y
Plots a single set of XY pairs connecting symbols with vertical and horizontal lines, starting with horizontal. Format data columns as: XY Pairs Single X Single Y
Multiple Horizontal Step Plot
Plots a multiple sets of XY pairs connecting symbols with vertical and horizontal lines, starting with horizontal. Format data columns as: XY Pairs Many X Many Y Single Y, Many X Single X, Many Y
Area Plots
Simple Area
Plots single set of XY pairs as a line plot with a downward fills. Format data columns as: XY Pairs Single X Single Y
126 Chapter 4
Multiple Area
Plots multiple sets of XY pairs as line plots with downward fills. Format data columns as: XY Pairs Many Y Single X, Many Y Many X Single Y, Many X
Vertical Area
Plots single set of YX pairs as a line plot with a left direction fill. Format data columns as: Single X YX Pair
Multiple Vertical Area
Plots multiple sets of YX pairs as line plots with left direction fills. Format data columns as: Many X Single Y, Many X
Complex Area Plot
Plots multiple line plots with downward fills and intersections. Format data columns as: XY Pairs X Many Y Y Many X Many X Many Y
Polar Plots
Scatter
Plots angle and distance data as symbols. Format data columns as: Theta, R Pairs XY Pairs Many Theta Many R Single Theta, Many R R, Many Theta
Lines
Plots angle and distance data points connected with lines. Format data columns as: Theta, R Pairs XY Pairs Many Theta
128 Chapter 4
Plots angle and distance data as symbols connected with lines. Format data columns as: Theta, R Pairs XY Pairs Many Theta Many R Single Theta, Many R R, Many Theta
Ternary Plots
Scatter
Plots ternary triplet data as symbols. Format data columns as: Ternary Triplets Ternary XY Pairs Ternary YZ Pairs Ternary XZ Pairs
Lines
Plots ternary triplet data as data points connected with lines. Format data columns as: Ternary Triplets Ternary XY Pairs Ternary YZ Pairs Ternary XZ Pairs
Scatter & Lines
Plots ternary triplet data as symbols connected with lines. Format data columns as: X,Y, and Z values; or data. Ternary Triplets Ternary XY Pairs Ternary YZ Pairs Ternary XZ Pairs
Plots a single column of data as Y values. Format data columns as: XY Pair Single Y
Grouped Bar
Plots multiple columns of data in a series of bars. Format data columns as: Single X, Many Y
130 Chapter 4
Plots data as Y values with error bars. If using worksheet columns or asymmetric error bar columns as the symbol value source, format data columns as: Single Y XY Pair If using columns means, first column entry, or last column entry as symbol values, format data columns as: Single X Many Y Many Y If using row means, row median, first row entry, or last row entry, format data columns as: Single Y Replicate X, Y Replicate
Grouped Error Bars
Plots data as multiple sets of Y values in a series of bars with error bars. If using worksheet columns or asymmetric error bar columns as the symbol value source, format data columns as: Many Y Single X, Many Y
If using row means, row median, first row entry, or last row entry, format data columns as: Many Y Replicates Single X Many Y Replicates Error bar values are from the worksheet.
Stacked Bars
Plots multiple columns of data as a series of stacks in bars. Format data columns as: Single X, Many Y Many Y Many Y Replicates Single X, Many Y Replicates
Plots a single column of data as X values. Format data columns as: XY Pairs Single X
Grouped Bar
Plots multiple columns of data in a series of bars. Format data columns as:
132 Chapter 4
Plots data as X values with error bars. If using worksheet columns or asymmetric error bar columns as the symbol value source, format data columns as: Single X YX pair If using columns means, first column entry, or last column entry as symbol values, format data columns as: Many X; Single Y, Many X If using row means, row median, first row entry, or last row entry, format data columns as: Many X Replicates Single Y, Many X Replicates Error bar values are from the worksheet.
Grouped Error Bars
Plots data as multiple sets of X values in a series of bars with error bars. If using worksheet columns or asymmetric error bar columns as the symbol value source, format data columns as: Single Y Many X
Many X If using row means, row median, first row entry, or last row entry, format data columns as: Many X Replicates Single Y, Many X Replicates Error bar values are from the worksheet.
Stacked Bars
Plots multiple columns of data as a series of stacks in bars. Format data columns as: Single Y, Many X Many X Single Y Many X Replicates
Box Plots
Vertical
Plots the median, 10th, 25th, 75th, and 90th percentiles as vertical boxes with error bars. Format data columns as: Many Y Single X, Many Y Error bar values are column means.
134 Chapter 4
Horizontal
Plots the median, 10th, 25th, 75th, and 90th percentiles as horizontal boxes with error bars. Format data columns as: Many X Single Y, Many X Error bar values are column means.
Contour Plots
Contour
Plots data XYZ values in 2D space. Format data columns as: XYZ Triplet Many Z XY, Many Z
Filled Contour
Plots data XYZ values in 2D space filling in the area between contour levels. Format data columns as: XYZ Triplet Many Z XY, Many Z
3D Line Plots
3D Trajectory
Plots data as XYZ data points, but only displays X or Y gridlines. Format data as: Many Z Single XY Many Z
136 Chapter 4
Figure 4-3 In this worksheet, the data is arranged for an XY Categories data format. The "Animals" column is what you would select as the "category" column in the Graph Wizard.
The Category Data format is available when creating summary plots. Graph types and styles that can use a category data format are: Scatter Plot. Multiple Scatter; Multiple Regression Line Plot. Multiple Straight Lines; Multiple Spline Curves; Multiple Vertical Step Plot; Multiple Horizontal Step Plot; Multiple Vertical Midpoint Step Plot; Multiple Horizontal Midpoint Step Plot For more information, see Plotting Category and Grouped Data in Chapter 6.
138 Chapter 4
Using the Same Column for Multiple Curves (Single X or Y vs. Many Y or X)
SigmaPlot can graph many curves using the same X or Y data column. There is no need to duplicate a column that is used for more than one curve; for example, enter the X data into only one column, and enter the corresponding Y data into as many columns as you have curves. Order and length of columns does not matter.
Figure 4-6 Data for a 2D Graph Arranged and Picked as X Many Y
140 Chapter 4
Arranging Data Using Column Means. Plots the average of an entire worksheet column as a single data point, then uses the column statistics to compute error bars, as specified by the Error Calculation. Arranging Data Using the Column Median. Plots the median of an entire worksheet column as a single data point, then uses the column statistics to compute error bars, as specified by the Error Calculation.
Its corresponding worksheet has three stack level value columns. Each column contains four rows.
The X values are in the first column. The second column, Stack Level 1, contains the bar heights for the first stack level. This appears as the first bar in each stack, in black. The bar heights for the second and third stacks appear the third and fourth columns, Stack Level 2 and Stack Level 3. The values in the Y rows are the bar heights for each stack. In this example, because the values for each row (Stack Level 1 through Stack Level 3) add up to 100, the height of each stack in the graph will also be 100.
142 Chapter 4
and
where R is the radius, and q is the angle of the data point from the origin.
144 Chapter 4
146 Chapter 4
2. Type the transform function as follows: pi=3.14159265359 col(diameter)=sqrt(col(area)*factor/pi) where diameter is the column number for your diameter data, area is the column number for your original data to be represented by area, and factor is some number to increase or decrease the magnitude of your data to a reasonable range. Tip: Reduce the diameters of your symbols to a reasonable size before plotting them. 3. Click Run. Your new data appears in the worksheet. If you change the symbol shape, you must use a different equation to transform area data.
Note: If multiple Z columns are plotted, they all must be next to each other. The X and Y columns can be located anywhere.
Y1
Y2
Y3
Y4
X1 X2 X3
Z1 Z2 Z3
Z4 Z5 Z6
Z7 Z8 Z9
The arrangement of this data for the three possible methods of picking columns to plot are described in the following sections. X, Y, and Z Data in Three Columns. To plot three columns as the X, Y, and Z values of a contour or mesh plot, the data must be in long form mesh format. This format assigns the proper Z value to each X and Y point in the mesh, in the required order.
148 Chapter 4
For example, for the table of X, Y, and Z values shown above, the three column mesh format must be arranged in the worksheet as:
Table 3-2
X data
Y data
Z data
X1 X2 X3 X1 X2 X3 X1 X2 X3 X1 X2 X3
Y1 Y1 Y1 Y2 Y2 Y2 Y3 Y3 Y3 Y4 Y4 Y4
This arrangement places the XYZ data point coordinate values in the required order. The XYZ columns must be the same length.
Figure 4-10 Data Arranged in Long Form Mesh Format
X and Y Columns vs. Many Z Columns. You can also place the X and Y data in single columns, then place the corresponding Z data in many continuous columns. This method may work best if you have XYZ data displayed in a table, or if you have irregularly incremented X or Y values.
To use this option, you should have as many Z columns as you have Y rows, and the Z columns should be the same length as the X column.
Table 3-3
X data
Y data
Z data
X1 X2 X3
Y1 Y2 Y3 Y4
Z1 Z2 Z3
Z4 Z5 Z6
Z7 Z8 Z9
The data in the first Z column is assigned to the first Y value, the data in the second Z column to the second Y value, etc. The data in each row of the X column is assigned as the X value for the data in the same row in the Z columns.
Figure 4-11 XYZ Data Arranged as One X Column, One Y Column, and Many Z Columns
The X and Y data must be strictly ascending or descending. Note that in this case, you can use columns of uneven length. Extra X, Y, or Z values created by uneven columns are not plotted, as mesh plots cannot graph missing values. Z Data vs. Row and Column Numbers: You can also plot columns as Z values versus the cell columns and row numbers as the X and Y values. This is the appropriate column assignment option to use: for mesh plots and 3D Bar Charts where X and Y values are evenly and equally spaced; for example, when graphing pixel intensity data for an image. All data is assigned as a Z value, and the Z columns must be contiguous. To use this format for a mesh plot, no special data arrangement is required other than equal column length. The rows and columns of the cells can be used as either the X or Y values.
150 Chapter 4
Creating Graphs
The most typical way to create a graph in SigmaPlot is with the Graph Wizard. You can start the Graph Wizard either by: Clicking the Graph Wizard button on the Standard toolbar. Clicking Create Graph on the Graph menu. Clicking any graph type on one of the Graph toolbars. Using the Graph Style Gallery. You can also create graphs using graph page templates. For more information, see Using Graph Pages as Templates in Chapter 5. Then follow the instructions as they appear in the Graph Wizard and click Next to move to the next panel. Tip: You can either select the worksheet columns to plot before creating your graph by dragging the pointer over your data, or you can select data columns later in the Graph Wizard. You can even select data ranges. For more information, see Entering Data Ranges into the Graph Wizard below.
You can also create pre-formatted worksheets. These are empty worksheets that contain predefined sections that represent the data formats used for creating SigmaPlot graphs. To do this, click Create a Graph. From here, you have all the control of creating graphs that you have in the standard SigmaPlot Graph Wizard, but in one dialog box. 1. Select a graph type. On the top of the dialog box, you see a row of Graph Style icons. Click one. Its available graph styles appear on the Graph Style menu on the left hand side of the Start Up screen.
2. Select a graph style. Depending on the Graph Style you choose, all its available data formats appear in the Type drop-down list under Formatted worksheet. A preview of what the worksheet will look like appears on the right. A preview of the graph appears directly above it. Options that are not available for a particular graph type remain grayed out. 3. Click OK and enter the data. Once you click OK, an empty graph page appears along with an empty formatted worksheet. Now you can enter or import (even copy and paste!) the data into the formatted worksheet. Of course, this is assuming the data you are importing or copying and pasting uses the same data format. As you enter the data, you will see the graph grow as it is being created. For more information, see Entering Data into a Worksheet in Chapter 3. 4. Modify the data. All you have to do to modify the data is select a cell in the worksheet and enter the data point again. The graph will change simultaneously. 5. Modify the graph. Just double-click the graph, and the Graph Properties dialog box appears. For more information, see Modifying Graphs below. 6. Create another pre-formatted worksheet. At this time, the only way to create a new preformatted worksheet is to close and restart SigmaPlot.
152 Chapter 4
1. On the Standard toolbar, click the Graph Wizard button. The Graph Wizard appears.
Figure 4-12 Graph Wizard Graph Types
2. Under Graph Types, select the type of graph you want to make. 3. Click Next. 4. Under Graph Styles, select the desired graph style.
Figure 4-13 Graph Wizard - Style
5. Click Next. If the graph style you have chosen uses error bars, you are prompted to choose an error bar source and a value to use for the error bars. For more information, seeCreating 2D Scatter Plots with Error Bars in Chapter 6.
6. Click Next. 7. Under Data format, select how your data is formatted, and click Next.
Figure 4-14 Specifying the Data Format
8. From the Data for drop down list, select the worksheet columns that correspond to the axis or error bar of your plot. You can also drag a range of data on the worksheet using the mouse. Note: When creating graphs using Microsoft Excel, you can only enter ranges manually. You can also select a range of data by entering the range manually into the Data for box. After entering the range, press Enter. The range appears in the Graph Wizard. For more information, see Entering Data Ranges into the Graph Wizard below. If you make a mistake while selecting data, double-click the mistaken column in the Selected Columns list to clear the selection.
154 Chapter 4
For more information, see Creating SigmaPlot Graphs Using MicroSoft Excel below. 9. Click Finish to create the plot.
Excel Notation. You can also use Excel notation in the Graph Wizard. In Excel notation, the columns are alphabetized in lexicographic order and the rows are numbered. In this case, to specify a rectangular region, you would again specify the upper left and lower right cells. For example, both A3:D9 and $A3;$D9 specify a region with the upper left cell in the first column, third row and the lower right cell as the fourth column, ninth row. Note that the separator is a colon. The letters are case insensitive.
Figure 4-17 Selecting a Range of Data Using the Excel Format
Column Numbers Notation. You can make a selection of a consecutive group of entire columns by specifying the range of column indices. For example, to specify columns 1 through nine, type 1:9 or 1..9.
156 Chapter 4
Figure 4-18 Selecting a Range of Data Using the Column Numbers Format
Creating Graph Style Gallery graphs from the Graph Wizard. For more information, see Creating Graph Style Gallery Graphs from the Graph Wizard below.
1. Open the graph that you wish to add to the Graph Style Gallery. 2. If the Graph Style Gallery is not visible on your SigmaPlot desktop, from the menus select:
View Graph Style Gallery
3. From the graph page, select the graph and drag and drop it into the Graph Style Gallery window. A thumbnail of the graph appears in the Graph Style Gallery palette. The graph title appears as the graph styles name.
To use the right-click short cut menu:
1. Select the graph on the page. 2. Right-click and on the shortcut menu click Add Graph. The graph style appears in the Gallery.
158 Chapter 4
Figure 4-19 Using the Right-Click Shortcut menu to Add a Graph to the Graph Style Gallery. Here, the Graph Style Gallery is docked.
2. Double-click the graph style you want to use. The Graph Wizard - Create Graph panel appears. For more information, see Creating Graphs Using the Graph Wizard above. 3. Select the worksheet columns you want to use for the plot. 4. Click Finish to create the plot.
160 Chapter 4
3. Double-click to open the Graph Properties dialog box, or use any of SigmaPlots editing tools. For more information, see Modifying Graphs below. 4. Close the graph page. The new style is saved in the Graph Style Gallery.
1. On the Standard toolbar, click the Graph Wizard button. The Create Graph - Type panel of the Graph Wizard appears.
Figure 4-21 You can choose Graph Gallery as a Graph Type when creating graphs using the Graph Wizard.
2. Under Graph Types, select Graph Gallery, and click Next. The Create Graph - Gallery panel of the Graph Wizard appears. All graphs that appear in the Gallery graphs list are also in the Graph Styles Gallery.
3. Under Gallery graphs, select the graph type that you want to apply to your data, and click Next. The Create Graph - Select Data panel of the Graph Wizard appears.
Figure 4-23 Selecting Data in the Graph Wizard
4. Under Data for, select the worksheet columns to plot. If you make a mistake while selecting data, select the correct column in the Selected Columns list. 5. Click Finish to create the graph. A graph appears on the page using the applied Gallery graph style.
162 Chapter 4
Modifying Graphs
Use the Graph Properties dialog box to make most graph modifications. To display the Graph Properties dialog box, double-click the graph.
To hide or show graph titles and automatic legends, to hide or show plots, and to make modifications to automatic legends, click the Graph tab, and under Settings for, click Legends. To apply your changes, click Apply, or click OK to apply your changes and close the Graph Properties dialog box.
3. Place the pointer over the desired graph or plot and click.
164 Chapter 4
Naming Plots
The default plot names are numeric; for example, Plot 1, Plot 2, etc.
To assign a new name to a plot:
1. On the Standard toolbar, click the Graph Properties button. The Graph Properties dialog box appears.
Figure 4-27 Using the Graph Properties dialog box to rename a graph. Click Rename to open the Rename dialog box.
2. Click the Plots tab. 3. From the Plot drop-down list, select the plot to rename. 4. Click Rename. The Rename Item dialog box appears.
Figure 4-28 Type a new name for the plot in the Rename Item dialog box.
5. Type a new name. 6. Click OK. The Rename Item dialog box closes.
166 Chapter 4
Naming Graphs
The default graph names are numeric, and include the graph type; for example, 2D Graph 1, 2D Graph 2, and so on.
To assign a new name to a graph:
1. Double-click the graph title that appears above the graph to select it. 2. Type the new name, making any font changes as necessary using the Format Text toolbar. 3. Click elsewhere on the graph when finished.
Figure 4-29 The Graph Wizard displays the available Data Formats for the current plot
3. Under Data Format, select a data format, and click Next. 4. If you do not change the data format for your graph, your previous column choices appear under Selected Columns. To change column assignments, under Selected Columns, select the desired assignment, then under Data For, select the appropriate column from the worksheet or from the data list.
Figure 4-30 You can change the column assignments using the Graph Wizard.
Note: To clear a column assignment by double-click it in the Selected Columns list. 5. If you change the data format for your graph, a single data type is highlighted in the Selected Columns list. To pick data, either click the corresponding column directly in
168 Chapter 4
the worksheet, or choose the appropriate column from the Data for list. Use this method to pick X, Y, or Z data, R and theta data, and error bar data, if applicable. 6. If you make a mistake while picking data, click the mistaken entry in the Graph Wizard, then choose the correct column from the worksheet. 7. Repeat the process for every data column. When you have chosen the data appropriate for your style of plot, click Back to repick data columns, or if applicable, click Next to pick data for additional plots. 8. Click Finish to close the Graph Wizard and view the changed graph.
1. Click the plot to modify. Square handles appear over the data points for the clicked curve. Do not click the graph, or you will add a plot to the graph. 2. On the Standard toolbar, click the Graph Wizard button. The Graph Wizard appears displaying the data format of the current plot. 3. To change plot style, click Back to view the Graph Styles list. Choose from the list of available styles then click Next.
Figure 4-31 You can use Graph Wizard to change the type and style of the graph.
4. To change the plot type, click Back twice to view the Graph Types list. Choose from the list of available graph types, then click Next. 5. Click Next until you can select a data format again for the new plot type or style from the Data Format list, then click Next. You are prompted to specify which worksheet columns to plot. 6. If necessary, repick the data columns to plot. Otherwise, click Finish to complete you plot type or style change.
170 Chapter 4
Figure 4-32 In this example of a graph with two plots, each plot has separate Y-axes
1. Click the graph to modify. Note: Small square handles surround the graph. Do not click a curve, or you will modify that curve instead. 2. From the menus select:
Graph Add Plot
The Graph Wizard appears displaying all the graph types. The available styles and types for a new plot are limited depending on the other plot types and styles in the current graph; for example, you cannot add a Polar plot to a 2D Cartesian plot, or vice versa.
Note: If the selected graph cannot accommodate the plot type or style that you want to add, the plot will be created as a new graph. You can move the graph of the new plot over the original graph so that it appears to be in the same graph. 3. Select a Graph Type and click Next. 4. Select a Graph Style and click Next. 5. Select a Data Format and click Next. 6. Pick data either by clicking the corresponding column directly in the worksheet, or choosing the appropriate column from the data list. Use this method to pick X, Y, or Z data, R and theta data, and error bar data. Note: If you make a mistake while picking data, click the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list. 7. Repeat the process for every data column. When you have chosen the data appropriate for your style of plot, click Back to re-pick data columns, or if applicable, click Next to pick data for additional plots. 8. Click Finish.
1. Right-click the plot. 2. On the shortcut menu, click Hide. The plot is hidden, but not removed.
172 Chapter 4
Figure 4-33 You can use the right-click shorcut menu to hide graphs.
2. Click the Graph tab. 3. Under Settings for, click Plots. All plots associated with the current graph are listed under Show/hide plots. A check mark in the check box next to the name of a plot indicates that the plot is displayed. 4. Clear a check box to hide a plot from view, or select it to show the plot.
To delete a plot:
3. Choose the plot you want to delete. 4. To delete the individual curves of a plot, select a curve on a graph, then press the Delete key.
1. Double-click the graph. The Graph Properties dialog box appears. 2. Click the Plots tab. 3. Select the desired plot from the Plot drop-down list.
174 Chapter 4
4. To plot only a portion of your data, under Data sampling, select Only Rows, and then enter the range to plot. 5. To sample the column rows by a specified increment, select by and type a number. Typing two samples every other row reduces the number of rows plotted by 50%, typing three samples every third row, and so on. You can also use the By list to select a number of rows plotted.
Figure 4-35 You can sample data using the Plots tab on the Graph Properties dialog box.
The graph on the left plots both a missing data point and out-of-range data point. The graph on the right ignores both missing and out of range points.
To ignore missing and out of-range points:
176 Chapter 4
Figure 4-37 Graph Properties Dialog Box Plots Tab Data Settings
2. Click the Plots tab. 3. Select Data from the Settings for list. 4. Select the desired plot from the Plot drop-down list. 5. To plot data without missing values, under Ignore, select Missing values. To plot missing values, clear the option. 6. To plot data without out of range values, under Ignore, select Out of Range Values. To plot out of range values, clear the option. 7. Click OK.
178 Chapter 4
1. On the Excel toolbar, click the SPW button, or on the Excel Insert menu, click SigmaPlot graph. The Graph Wizard appears. 2. Select Excel data and create the graph using the Graph Wizard.
180 Chapter 4
Figure 4-38 Graph Properties Dialog Box Plots Tab Symbols Setting
2. Click the Plots tab. 3. From the Settings for list, select Symbols. 4. From the Plot drop-down list, select the plot to modify. 5. To change the symbol type for the selected plot, from the Type drop-down list select a symbol type, or choose to increment symbols using the one of the symbol schemes. To create a plot that displays lines only, turn off symbols by choosing (none). For more information, see Automatically Incrementing Symbols below. 6. To change the size of the symbol, move the Size slider, or type a new value in the Size box. By default, all symbols in a plot are the same size. Use symbols of different sizes by entering symbol sizes in a worksheet column, then selecting the column from the Size list. 7. To change the fill color of symbols for the selected plot, under Fill Color, select a color from the Color list, or choose to increment fill colors using one of the incrementing
schemes. To turn off symbol fills, select (none). For more information, see Using Custom Symbol, Fill, Line, and Color Increments below. Select (Custom) to open the Color dialog box to create or choose a custom color. For more information, see Using Custom Colors in Chapter 5. Note: Hollow Symbols are symbols that use (none) as the fill color. They are hollow, that is, they are composed of the edge lines only. Lines, error bars, and graph background colors all show through unfilled symbols. This is useful if you have many overlapping data points. 8. To change the edge color of symbols, from the Edge Color drop-down list, select a color, or select to increment edge colors using one of the incrementing schemes. To turn off symbol edge color, select (none). Use the (Custom) option to open the Color dialog box from which you can create or choose a custom color. 9. To control the color of symbol dots and crosshairs, or of text used as symbols, use the Edge Color option. If a symbol is filled with black and has a black edge, then dots and crosshairs automatically default to white. 10. To change the thickness of the symbol edge, move the Thickness slider, or type a new value. 11. Click OK. For more information, see Using Characters and Text as Symbols below.
182 Chapter 4
Figure 4-39 Both graphs use the Doubles symbol scheme and the Black and White color scheme. The first graph has only one curve; the second has four.
Symbol types and colors appear on the curves of the plot in the same order as the symbol types and colors in the right-click popup menus of the incrementing option. For more information, see Using Custom Symbol, Fill, Line, and Color Increments below.
To automatically increment symbols:
1. Double-click the plot. The Graph Properties dialog box appears. 2. Click the Plots tab. 3. From the Settings for list, select Symbols. 4. From the Plot drop-down list, select the desired plot. 5. To increment symbol types and fill and edge colors automatically, under Symbols, from the Type, Fill Color, and Color lists, select a symbol scheme.
Figure 4-40 Right-click the symbol type to select the first symbol of the incrementing scheme.
Note: Increment schemes do not include (None) as a symbol type. 6. To change the first symbol type or color used in the incrementing sequence, from the Symbols Type, Fill Color, and Edge Color drop-down lists, select Incrementing. Rightclick the selected Incrementing option, and from the shortcut menu, click First Symbol or First Color, then click the symbol type or color to start the incrementing sequence. 7. Click OK.
184 Chapter 4
1. Enter the text you want to use as symbols in a worksheet column in the order you want the curve(s) to use them. To use numeric values as symbols, add a space after each value in the worksheet. You can assign the numbers that appear aligned to the left as symbols.
Figure 4-42 Example of Worksheet with Plot Symbol Text Entered in Column 3
You can use all the non-keyboard characters available for the default font. To view and access these characters, you can use the Windows Character Map utility. The Windows Users Guide also lists these special characters, along with the keystrokes required to enter them. 2. On the Standard toolbar, click the View Page button. 3. Double-click the plot on which you want to use text symbols. The Graph Properties dialog box appears.
186 Chapter 4
Figure 4-43 Change the font for text symbols by right-clicking the Type option and choosing Symbol Font.
4. Click the Plots tab. 5. From the Settings for list, click Symbols. 6. Under Symbols, from the Type drop-down list, select the column that contains the text or numeric values you want to use as symbols. Note: The column option does not appear in the Type list unless text or symbols are entered in a worksheet column. 7. Under Symbols, right-click the Type box, and from the shortcut menu, click Symbol Font. The Text Properties dialog box appears.
Figure 4-44 Change the font for text symbols by right-clicking the Type option and choosing Symbol Font.
8. Click the Font tab. 9. Select another font from the Font drop-down list. This feature is especially useful if you wish to use Wingdings, Zapf Dingbats, or other iconic or symbolic fonts as a symbol. The Fill Color and Edge Thickness options do not apply to text and characters. 10. Click OK.
188 Chapter 4
Figure 4-45 Using Symbol Sizes from a Worksheet Column for Plot Symbols
Symbol sizes are assigned to symbols and curves (or points, if the plot has only one curve) in the same order as the column pairs that form the curves which are listed in Graph Wizard.
To use worksheet values for symbol size:
1. Select the first cell of an empty column in the worksheet containing data for the current plot. 2. Type the size values to use in the order you want to use them. Since the symbol sizes correspond to symbol diameters or widths, make sure that the symbol sizes you enter
are of a reasonable size, that is, small fractions of inches or only a few millimeters or points. If desired, you can also include the measurement unit for the value. For example, for inches type in, for millimeters type mm, or for points type pt.
Figure 4-46 Example of Worksheet with Symbol Sizes Entered in Column 3
If you omit the measurement unit, the numeric values in the symbol size column are assigned the measurement unit specified in the Options dialog box Page tab. 3. Click the toolbar button to view the graph page. 4. Double-click the plot. The Graph Properties dialog box appears. 5. Select the plot that contains the symbols to modify from the Plot drop-down list. 6. Use the Size drop-down list to choose the worksheet column containing the symbol size values.
190 Chapter 4
Figure 4-47 Using the Plots Tab to Select Symbol Size from a Worksheet Column
7. Click OK. When creating a bubble plot, the Graph Wizard automatically prompts you to pick a column to specify bubble size. For more information, see Bubble Plots in Chapter 6.
2. Click the Plots tab. 3. Select Lines from the Settings for list. 4. Under Line style, from the Type drop-down list, choose a line type. For more information, seeUsing Custom Symbol, Fill, Line, and Color Increments below. Tip: To create a plot that displays symbols only, choose (None) to turn off lines. 5. To change the thickness of the line, type the new value in the Type box. 6. Choose a line shape from the Shape drop-down list. 7. To change the color of the lines in the selected plot, select a color from the Color dropdown list, or choose to increment line color using one of the incrementing schemes. Select (None) to create transparent lines. This in effect turns them off. Use (Custom) to create or choose a custom color. For more information, see Using Custom Colors in Chapter 5.
192 Chapter 4
8. To change or add a gap color, select a color from the Gap color drop-down list. This option is only available if you select a line type with actual "gaps" in it, like dotted or dashed, for example. 9. To control the layering of plot lines, use the Layering drop-down list to place lines behind or in front of plot symbols. Note: Hollow symbols (None) will always show plot lines. 10. Click OK.
For more information, see Using Custom Symbol, Fill, Line, and Color Increments below.
2. Click the Plots tab. 3. Select Lines from the Settings for list. 4. Select a plot from the Plot drop-down list. 5. From the Type and Color drop-down lists, choose a line scheme. Note: Windows is limited in its ability to supply the true colors for lines by the number of system colors available. For the best representation of true line colors, set your display to either HiColor (16-bit) or TrueColor (24-bit). 6. Right-click the incrementing option selected in the Type and Color drop-down lists, and from the shortcut menu, select First Line or First Color. 7. Choose First Line or First Color from the shortcut menu.
194 Chapter 4
8. Choose the line type or color to start the incrementing sequence. 9. Use the Line Thickness, Shape, Line Color, and Layering options to modify the lines, if necessary. For more information, see Changing Plot Line Attributes above. 10. Click OK.
Figure 4-51 Example of a Bar Chart with a Gray Scale Fill Color Scheme
196 Chapter 4
2. Click the Plots tab. 3. From the Plot drop-down list, select the plot that contains the fills to modify. 4. From the Settings for list, select Fills. 5. To change the background fill color, under Fill Color, from the Color list, select a color, or choose to increment fill colors using the one of the incrementing schemes to change the background fill color. 6. To turn off background fills, select (None). 7. To create a custom color, select (Custom). For more information, see Using Custom Colors in Chapter 5. 8. To change the fill pattern and density for the selected plot, under Pattern and Edge, from the Pattern list, select a fill pattern, or select to increment fill patterns using one of the fill schemes. To turn off fill patterns, select (None). 9. To change the thickness of the pattern lines and edges, move the Thickness slider.
198 Chapter 4
3. From the Plot drop-down list, select the plot that contains the fills to modify. 4. From the Settings For list, select Fills. 5. Select a scheme from the Color and Pattern drop-down lists. Colors and patterns appear in the bars, boxes, or pie chart slices of the plot in the same order as the right-click shortcut menu. 6. Right-click the incrementing option and from the shortcut menu, select First Pattern or First Color, and then select the pattern or color to start the incrementing sequence. 7. Click OK. For more information, see Changing Plot Fill Patterns and Colors above.
200 Chapter 4
Figure 4-55 Using the Insert Graphic Cells Dialog Box to Specify a Custom Line Sequence
3. Click the Colors, Lines, Symbols, or Patterns tab. Note: Using symbol types from a column specifies the symbol shape only. If you want to change the symbol fills, create another color column and use it as the symbol fill colors. Typically, white is used for hollow symbols, and black for solid symbols. 4. Select the first cell in an empty column in the worksheet. 5. Double-click the color, line, symbol, or fill pattern in the Insert Graphic Cells dialog box you want to place in the cell. Note: Do not mix graphic cell types within the same column; for example, place colors in one column, symbols in a different column, fills in yet another column, and lines in a fourth column. However, you can use multiple columns to define several different increments of the same graphic cell type. For example, you can have several columns containing colors of differently ordered increments. The item appears in the worksheet cell. 6. Continue adding to the column, in the order you want the curves to use the colors, lines, symbols, or patterns. The order of the curves is the order in which they appear in the Selected Columns drop-down list in the Graph Wizard.
7. Close the Insert Graphic Cells dialog box. 8. Click the View Page button. 9. Double-click the plot. The Graph Properties dialog box appears.
Figure 4-56 Assigning Custom Symbol Colors in a Worksheet Column to a Plot
10. From the Plot drop-down list, select the plot to modify. 11. From the Settings for list, select Fills, Area Fills, Symbols, or Lines, depending on what you have defined in the worksheet. 12. Choose the name of the column which contains the appropriate graphic cells from the Symbols Type, Fills Foreground Pattern, or Lines Type, or Color drop-down lists. If you are applying a large number of colors or other property schemes, you may wish to turn off the automatic legend, which will attempt to display your first 25 different data points. For more information, see Editing Automatic Legends in Chapter 5. 13. Click OK.
202 Chapter 4
To control bar and box width and spacing for bar charts and box plots:
1. Double-click the plot to modify. The Graph Properties dialog box appears. 2. Click the Plots tab. 3. From the Settings for list, select Widths. 4. To change the width and spacing between bars for all bar charts and box plots, move the Bar thickness slider. The wider the bars or boxes, the less space between them. The narrower the bars or boxes, the more space between them. 5. To change the width and spacing between groups of 2D and 3D bars, move the Group spacing slider. This option is only available for grouped and 3D bar charts. SigmaPlot sets grouped bar widths and spacing to as wide or as narrow, and as far or as close as possible, given the corresponding spacing or width setting. 6. To set a constant width for all bars or boxes, from the Width drop-down list, select Uniform. This is the default setting. If the bars are set to Uniform, the Bar thickness
setting has the same effect on all bars. For more information, see Uniform versus Variable Bar Widths below. 7. To set potentially uneven widths for bars and boxes, select from the Width drop-down list, select Variable. If the constant column values are uneven, the bars will vary in width according to the corresponding axis values. Change bar widths according to the percent of their total widths, if the bars are set to Variable, so that wide bars are more affected than thin bars. Note: Bars created with a single plot will not overlap. However, you can create bars using separate plots and overlap them. For more information, see Spacing Bars from Different Plots in Chapter 6. 8. To create a needle plot, move the Bar thickness slider to set bar widths to the narrowest possible widths.
204 Chapter 4
Figure 4-58 To make a histogram needle plot, create a bar chart and set the Bar Thickness to Needle.
9. To change bar alignment, from the Align drop-down list, select either Center, Left, or Right. By default, bar chart bars are centered around the data point. Use Align to alternately draw the bars right or left aligned with the data points.
Figure 4-59 From Left To Right: Bar Charts with Alignments to the Left of the X Points, to the Right of the X Points, and Centered over the X Data Points
206 Chapter 4
Figure 4-61 The graphs on the left are examples of 2D plots with drop lines to the Y and X axes. The graph on the right is an example of a 3D graph with drop lines to all axes.
Drop lines always fall towards the minimum of a range; for example, if a Y axis range were reversed, a drop line to the X axis would fall to the top of the graph rather than the bottom. Use the Drop Lines settings in the Graph Properties dialog box Plots tab to create new drop lines, and to modify existing drop line type, thickness, and color.
To add or modify drop lines for a selected plot:
1. Double-click the plot to modify. The Graph Properties dialog box appears.
Figure 4-62 Graph Properties Dialog Box Plots Tab Drop Lines Setting for a 2D Scatter Plot
2. Click the Plots tab. 3. From the Plot drop-down list, select the plot that contains the drop lines to modify. 4. From the Settings for list, select Drop Lines. 5. Select the X or Y drop-line check box. Drop lines are added to any and all planes or axes that are selected. 6. From the Type drop-down list specify the type of line to use for selected drop lines. 7. To adjust line thickness, type the new value in the Thicknessbox. 8. To set drop line color, select a color from the Color drop-down lists. Select any of the listed colors, or select (Custom) to select or define a custom color. For more information, see Using Custom Colors in Chapter 5. 9. Click OK.
208 Chapter 4
The Plot Equations dialog box Equation tab appears, either with Untitled or the name of the last used equation in the Name field.
Figure 4-64 Plot Equation Dialog Box Equation Tab and Functions Palette
2. To manually enter the equation, from the Name drop-down list, select Untitled. 3. If necessary, delete the existing equation in the f = field, and then either type the equation, or click the Functions Palette button to open the Functions Palette. The Functions Palette provides immediate access to some of the most frequently used functions. You can also select one of the last ten used functions from the Name drop-down list. For more information, see Plotting Saved Equations below.
210 Chapter 4
4. From the Variables group box, select either 2D or 3D. 5. Set the independent variables using the Name, Minimum, Maximum, and Intervals boxes. Name. Type the name of the independent variable(s). Minimum and Maximum. Type the extent of the range of values for the corresponding independent variables. Intervals. Set the number of intervals for sampling independent variables over a specified range. Note: You can also select a column in the worksheet. The range of that column appears in the Minimum and Maximum edit boxes. 6. To set the equation parameters, click the Options tab. For more information, see Setting Equation Parameters below. 7. Click Add As. The Add As dialog box appears.
Figure 4-65 Add As Dialog Box
8. Type the name of the equation in the Equation Name edit box. 9. Click OK. The equation name appears in the Name drop-down list on the Equation tab.
10. Click Plot. A graph page appears with the plotted equation, and the equation values appear in the worksheet. 11. Click Close to close the dialog box. If desired, you can add plot an equation and add it to the existing graph, or plot a new equation on a new graph page.
212 Chapter 4
3. Either manually enter the equation in the f = edit box, or choose an existing equation, or use the same equation as used previously if you want to change the parameters. 4. To set the equation parameters, click the Options tab. For more information, see Setting Equation Parameters below.
Figure 4-68
5. If you do not want to create a second graph page, select Add to current graph and clear Create new graph. 6. Click ThePlot, the plot appears on the current graph. 7. Click Close to close the Plot Equations dialog box.
214 Chapter 4
3. In the Parameters box, enter or edit the parameters. Enter parameters with the name of the parameter first, followed by an = sign, and then the value, i.e. a=3 or b=7.231 4. To assign a value to the next parameter, press Enter. 5. Click Plot to plot the equation.
2. Click the Library tab. 3. Select an equation category from the Equation category drop-down list. The items that appear in the Equation category drop-down list are sections in the Standard.jfl library. Below, in the Equation Name list, are items that appear under that section name in the notebook. 4. Select an equation from the Equation name list. 5. Click Select. The Equation tab appears with the selected equation displayed in the Name drop-down list.
Figure 4-71 Plot Equation Dialog Box Equation Tab
Some of the settings for SigmaPlots built-in equations in the Standard.jfl library are read-only. To modify a built-in equation, click Add As to create an equation based on the built-in equation. 6. Click Plot. A graph page appears with the plotted equation, and the equation values appear in the worksheet. 7. Click Close to close the Plot Equation dialog box.
216 Chapter 4
Solving Equations
Use the Equation Solver on the Plot Equations dialog box to evaluate mathematical expressions for functions and to solve equations. The Equation Solver uses the expression entered in the Equation tab on the Plot Equations dialog box as the basis for its results. This expression then appears on the Solve tab for evaluation.
To solve an equation:
2. Click the Equation tab, and enter an equation in the f = box. You can also select one of the last ten used functions from the Name drop-down list, or you can choose any of the built-in parameterized equations used by the Regression Wizard. Select these equations from the Library, too. For more information, see Plotting Saved Equations above.
3. Click the Solve tab. The entered equation appears in the f = box on the Solve tab.
Figure 4-73 Solve Tab of the Plot Equation Dialog Box
4. Under Options, select the mode of operation. You can select from one of the following: Evaluate F at. Enter a numerical value for each variable that occurs in the expression in the boxes that appear at the bottom of the dialog box. Solve equation for x within range. Enter a numerical value into the box which appears to the left of the expression (the default value is 0) to complete the definition of the equation. You must also enter limits for a range of values of the equation variable. The default range limits are taken from the values entered on the Equation tab. The Solver is only available for expressions containing a single independent variable, although any number of parameters can be present. 5. Under Options, click Evaluate or Solve, depending on the selected mode of operation. The resulting value or the equation solutions that lie between the prescribed ranges appear in the Results box.
218 Chapter 4
1. Using the Plot Equations dialog box, graph the function equation y = f(x). 2. Observe where the graph intersects the x-axis. This technique aids in determining range limits for the independent variable in the Function Solver (Solve tab of the Plot Equation dialog box). If the distance between two solutions of an equation is small relative to the size of the range, then the Function Solver may not return both solutions. The resolution of the solutions is approximately two orders of magnitude less than the size of the range. You can obtain higher resolution by adjusting the range limits to reduce the range size. There is particular difficulty, due to roundoff error, in determining solutions to 0 = f(x) at points where the graph of y = f(x) does not cross the x-axis, but lies on one side of it. An example of this situation is the graph of y = x^3+x^2 at x = 0. Although in many cases, as with the above equation, the Function Solver provides the solution, in some cases, however, the solution will not be found and recorded in the Results box. If you suspect that there is such a solution and the Function Solver does not find it, then try the following technique for approximating the solution: 3. Alter the value for the left side of the equation by a small amount. 4. Re-solve the equation. This is equivalent to slightly shifting the graph of the equation up or down until it lies on both sides of the axis. In general, the Results edit box then reports two solutions that are very close together. As smaller amounts are used to adjust the left side of the equation, these two solutions are seen to converge to one solution. As an example, try solving the equation 0 = sin(2*x)*cos(3*x) over the range from x = 1 to x = 2. The Function Solver will indicate that there are no solutions. Using the above technique will yield solutions that are close to the true solution of PI/2.
Spurious Solutions
A less frequent problem involves the appearance of spurious solutions. Due to the limits of floating point numbers, the value of an expression f(x) at x = a might compute to zero even if x = a is not a true solution to 0 = f(x). This situation commonly arises when the graph of y = f(x) is very "flat" near a point where it intersects the x-axis. For example, consider the equation 0 = x^201. If you solve this equation over the range from x=0 to x=1, then the Function Solver will return 13 solutions even though
220 Chapter 4
the only true solution is x = 0. This is because each of 13 results raised to the 201st power is equal to zero in the machines floating point representation.
3. Click the Regression Line tab. 4. Under Regressions, select either Each Curve to draw a regression for the data in each curve of the selected plot, or All data in plot to draw a single regression for all of the data in the selected plot from the Regressions group box. If neither box is selected, a regression is not drawn. If both boxes are selected, regressions are drawn for each curve and for all the data in the plot. 5. Under Line, select the desired regression order from the Order drop-down list. 6. Select the regression line type from Type drop-down list. 7. Select line color from the Color drop-down list. 8. To change line thickness, move the Thickness slider. 9. To set the extent of regression line(s) all the way across the graph, under Options, select Extend to Axes.
222 Chapter 4
.
Figure 4-76
Click Copy to copy the results and paste them into the worksheet, a report, or any other Windows application. For more information, see Linear Regression, Confidence, and Prediction Calculation below.
224 Chapter 4
2. Click the Confidence Intervals tab. 3. Choose the method of prediction to use from the Method drop-down list. Select either 95% or 99% for confidence and prediction intervals. 4. Select the Confidence Interval or Prediction Interval option and select a line type and color, then move the Thickness slider or enter a value in the Thickness box to set line thickness. Line color, type, and thickness options work identically to the regression line type, color, and thickness options. 5. Click OK.
is an n x 1 vector of residuals. The solution for the least squares estimates of the parameters is:
where X1 denotes the transpose of X. SigmaPlot uses the Cholesky decomposition to invert the X1Ymatrix. (see Dongarra, J.J., Bunch, J.R., Moler, C.B., and Stewart, G.W., Linpack Users Guide, SIAM, Philadelphia, 1979). This produces the regression curve:
For further details on matrix linear regression, refer to chapter 2 of Draper, Norman, and Smith, Harry, Applied Regression Analysis, Second Edition, John Wiley & Sons, Inc., New York, 1981. Confidence Interval Calculation. Given a set of n data points (xi, yi) from two columns in the worksheet, SigmaPlot computes the pth order polynomial regression:
where (b0, b1, ..., bp) are the p + 1 estimated parameters and 0 is the y value predicted for any X0. The confidence interval for this calculated regression is defined by the two confidence limits:
226 Chapter 4
and the t value for n - p - 1 degrees of freedom and the standard normal percentile equivalent z (z = 1.96 or 2.576 for 95% and 99% confidence intervals respectively) is computed from a six term rational polynomial approximation taken from Sahai, H. and Thompson, W., Comparisons of Approximation to the Percentile of t,2, and F Distributions, Journal of Statistical Computation and Simulation, 1974, Vol. 3, pp. 81-93. Prediction Interval Calculation. The prediction interval is calculated using the following equation:
228 Chapter 4
3. Select Reference from the Settings for list. 4. Select a reference line to draw by selecting its check box. You can add up to five lines for each plot. The default names and calculations are the names commonly used when employing reference lines for quality control charts. 5. To change the reference line name, select the line from the list, then edit the Label box for that line. 6. To display the label next to the reference line, select Left or Right for horizontal reference lines, or Top or Bottom for vertical reference lines. 7. To change the value or statistic used for the line, select an option from the Calc dropdown list. If you are not using a mean as the calculation, type a value to multiply the statistic by, or a value to use as a constant, in the box next to the Calc drop-down list. The calculation options apply only to the reference line highlighted in the Graph Properties dialog box list of reference lines. To set the reference line value to a specific value, select the Constant Calc option, and enter the value to the right. Automatically calculated statistics are derived from the plot data. All data points graphed, including multiple columns of data, are used for reference line calculations. 8. Use the Appearance options to set a line type, thickness, and color for the highlighted reference line. Each reference line can have separate line attributes. 9. Use the Direction drop-down list to draw reference lines horizontally or vertically. 10. Use the Layering drop-down list to draw reference lines either Behind or In Front of the selected plot. 11. Click Apply when finished modifying the current reference line, then highlight another reference line to continue modifying reference lines, or click OK.
Creating Histograms
Histograms are step, needle, or bar charts that represent counts of the data points that fall within specified ranges. The Histogram Wizard guides you through the steps in creating a histogram: generating frequency data, specifying the number of buckets or intervals, and selecting a graph style. The Histogram Wizard allows you to specify the number of bins into which you partition the source data. The range of each interval is identical; the total range is the data minimum to the data maximum. The number of bars, steps, or needles displayed is generally equal to the number of bins. You can also create a histogram with an uneven bucket size. For more information, see The Histogram Transform Function below.
1. Enter the data you want to analyze in an empty column of the active worksheet. 2. From the menus select:
Graph Histogram
The Histogram Wizard appears. 3. Select the data for the histogram by choosing the appropriate column from the Source data for histogram drop-down list.
230 Chapter 4
4. Select the column for the Output of histogram either from the drop-down list, or by clicking the column.
Figure 4-81 Selecting the Output for Bin Centers in the Histogram Wizard
5. Select the column for the Output for bin counts either from the drop-down list, or by clicking the column.
Figure 4-82 Selecting the Output for Bin Counts in the Histogram Wizard.
6. Click Next. The Histogram - Bin Options panel appears, with Automatic binning already selected. The algorithm calculates the number of bins for representation, based upon the number of data points. Approximate Bins = 3 + log10(N) * log10(N)/log10(2) where N = number of non-missing points.
Figure 4-83 The Histogram Wizard Buckets Dialog Box
7. To specify a different number of bins, clear Automatic binning and select a number from the Number of bins list. You can enter values from 1 to 100.
232 Chapter 4
8. Click Next. 9. Select a graph style from the Graph Styles list. A preview of the graph appears.
Figure 4-84 The Histogram Wizard Graph Style Dialog Box
Click Finish. The graph appears on the active graph page, or a new page if the worksheet has no associated graph pages. The X axis representing the buckets is titled Raw Data. The Y axis representing the frequency or the number of data points in each bin, is titled Bin Count. Both use a linear scale. Note: If you choose None, SigmaPlot displays the worksheet with the output column containing the histogram frequency data.
1. Enter the data to analyze in column 1 the bin values in column 2 of the worksheet. Bin values are used as the upper bounds (inclusive) of the histogram interval ranges. The number of data points that fall within each specified range is counted. The number of histogram bars is equal to the number of interval upper bounds entered. The number of values that fall beyond the largest upper bound is also counted. 2. From the menus select:
Transforms User-Define
The User-Defined Transform dialog box appears. 3. Enter the following transform into the Edit Transform box: col(3)=histogram(col(1),col(2))
234 Chapter 4
Figure 4-86 Graphing the results of the HISTOGRM.XFM transform as a bar chart
4. Click Run. The histogram data appears in column 3. 5. To graph the data, plot column 3 as a bar chart. For more information, see Creating 2D Plots in Chapter 6.
Chapter
Use Graph Pages to display and modify graphs that plot data from your worksheets. You can create as many graph pages as you wish per worksheet. New graph pages are associated with the current worksheet, and are placed in the current notebook section. This chapter covers: An overview of graph pages. For more information, see About Graph Pages below. Working with page objects. For more information, see Working with Page Objects below. Adding another graph to a page. For more information, see Adding Another Graph to a Page below. Zooming in and out. For more information, see Zooming In and Out below. Using graph pages as templates. For more information, see Using Graph Pages as Templates below. Cutting, copying, and pasting graphs and other objects. For more information, see Cutting, Copying and Pasting Graphs and other Page Objects below. Dragging and dropping graphs. For more information, see Dragging and Dropping Graphs below. Hiding and deleting objects from the page. For more information, see Hiding and Deleting Objects from the Page below. Modifying object colors and lines. For more information, see Modifying Object Colors and Lines below. Moving and sizing graphs and objects. For more information, see Moving and Sizing Graphs and Objects below.
235
Aligning page objects. For more information, see Aligning Page Objects below. Editing text. For more information, see Editing Text below. Working with automatic legends. For more information, see Working with Automatic Legends below. Changing graph page format. For more information, see Changing Graph Page Format below. Using custom colors. For more information, see Using Custom Colors below.
2. The Options dialog box appears. 3. Click the Page tab to set the graph page options.
1. Select and view the graph page. If you want to export specific graph(s), select the graphs you want to export to a file. 2. From the menus select:
File Export
The Export File dialog box appears. 3. Enter the file name, directory and drive for the export file destination. 4. Click Export. If you chose one of the graphic file formats, a secondary dialog box appears, asking you to enter some graphic format information.
Figure 5-1 Export Tagged Info File Dialog Box
5. Enter the desired DPI and Color Resolutions; for EPS files, these setting only affect the resolutions of the TIFF header, not the actual PostScript resolution. For metafiles, this setting affects only 3D graphs.
The higher the DPI and Color resolutions, the better quality the image, but also the larger the file. Limit the DPI and Color resolutions to the capability of the intended output device. For example, if you are going to create 600 dpi slide output, set the DPI resolution no larger than 600. 6. If you want to export only the selected graph(s) or objects, select Export selected only. 7. Click OK to create the exported file using the specified file name and graphic resolutions, if applicable.
1. Select and view the page window. 2. Click the Print button to print the page using all the default settings.
To set printing options before you print the graph page:
The Print dialog box appears. 2. Click Properties. The printer Properties dialog box appears. 3. Click OK when you are satisfied with the printer properties settings. The Properties dialog box closes. Note: The Properties dialog box options vary from printer to printer. 4. Click OK to print the report.
The Graph Wizard guides you through a series of dialog boxes to select the type and style of graph, and to select worksheet data for plotting. After you create the graph, you can open the Graph Wizard to add or modify plots and axes.
Graph Properties
The Graph Properties dialog box customizes the plots, axes, grids planes, titles and legends of your graph. Use it for more advanced modifications to your graph. To open the Graph Properties dialog box, double-click anywhere on the graph, or on the Graph menu, click Graph Properties. The Plots, Axes, and Graph tabs offer many customizing features. The tab that appears depends on where you click on the graph. Click the Help button to learn more about the specific options and controls for each tab.
Object Properties
The Object Properties dialog box modifies many graph attributes including drawn objects. Use the Object Properties dialog box to make simple modifications to the objects and graphs. The Line and Fill tabs change fill patterns, lines of your plots and objects. The Size and Position tab changes position, scaling and size for all selected objects. To open the Object Properties dialog box, select an object on the graph page, rightclick, and then on the shortcut menu, click Object Properties.
Text Properties
The Text Properties dialog box modifies font and paragraph text attributes for all text on a page. Use the Text Properties dialog box to change attributes of non-editable text, as well as attributes for multiple text labels, and making global text changes.
Selecting text properties with no selected text sets the default attributes for new text labels. To open the Text Properties dialog box, on the Format menu, click Text Properties.
You can edit, copy, paste, move, size and scale, delete or hide all selected page objects, including graphs, text, drawn objects, and pasted objects. The following table summarizes the results of selecting various objects on the graph page.
Select By Graphs Plots Axes Tick marks Tick labels Axis titles Legends Fills or Lines Opens Double-clickGraph Properties dialog box/Plots tab Double-clickGraph Properties dialog box/Plots tab Double-clickGraph Properties dialog box/Axes tab Double-clickGraph Properties dialog box/Axes tab Double-clickGraph Properties dialog box/Axes tab Double-clickEdit manually Double-clickEdit manually Right-clickGraph Properties dialog box/Plots tab
To use Alt-Click:
1. Click the object that you want to select, which may be covered by another object, and then press the Alt key. 2. While holding down the Alt key, repeatedly click the object until it is selected under the position that you initially clicked. As you repeatedly click you will cycle through all objects that overlay one another.
To use Alt-Arrow:
1. Click the object that you want to select and then press the Alt key. 2. While holding down the Alt key, repeatedly press one of the arrow keys (up, down, right, or left) to select the object of interest. Use of different arrow keys will cycle through the objects in a different order. Note: It is important that you click on top of the object that you eventually wish to select. For example, selecting the intersection of the horizontal and vertical grid lines slightly below the symbol will result in a different sequence of selectable objects, i.e. vertical grid line, right y-axis, horizontal grid line and plot error bars. It is not possible in this case to select the symbol. The figure below shows a graph with multiple overlaid objects. The symbol has been clicked and the vertical grid line has been selected. To select the symbol, for example, click the center of the symbol as shown in the figure. In this case the top object is the vertical grid line so it is selected. Repeated Alt-clicking will cycle through the four overlaid objects : vertical grid line, plot symbol, right y-axis and plot error bars. With the plot symbol selected, you might use the color panel on the Graph Toolbar to change the plot symbol color.
cursor is not over a data point you can read the position of the cursor using the x and y scales for that plot.
1. Click the Page tab. 2. Clear the option Show data values.
or Select a graph from one of the Graph toolbars, or Click the Graph Wizard button.
1. Select the graph you want to copy. 2. Press Ctrl+C. 3. Make the destination page the current page either by opening it, or if it is already open, select the graph page name from the Window menu. A check mark next to the page name indicates that it is the active window. Note: If the destination page is in a different notebook than the source page, you must close the source page, and any other open work in the source notebook. 4. Press Ctrl+V to paste the graph. The graph appears on the current page, and the graph data appears in the worksheet associated with the current page. Another method is dragging and dropping. For more information, see Dragging and Dropping Graphs below.
The page appears without any obstructions. 2. To return to normal view of the page, press any key on the keyboard. The screen returns to its normal appearance.
Figure 5-4 Using the Zoom Pointer to Select a Region on the Page
Use keyboard shortcuts while viewing the page window. The zoom keyboard shortcuts to view the page are: At 50% actual size, press Ctrl+5. At 100% actual size, press Ctrl+1. At 200% actual size, press Ctrl+2. At 400% actual size, press Ctrl+4 Entire page, press Ctrl+F Magnified for a specific region, press Ctrl+U.
2. Choose the desired zoom level to fit the page to the window, or to zoom to a full screen view. Select Custom and move the slider or enter a specific zoom level to set a percentage of magnification.
Figure 5-6 Graph Page Zomed to 200%
Note: Never use templates to add a graph to a page. Template pages are ordinary graph pages. Any graph page can act as a template page if it is copied to a section or used from the File menu New command to create a new page. All attributes from the page - size, color, margins, and orientation - are retained. Any graphs and other objects on the page are also duplicated. Template graphs automatically plot the worksheet column data that was selected when the graph was created. When applying a page to a worksheet, make sure your data is already arranged as required, or re-pick the data for the graph after applying the template. You can determine which columns are plotted by either looking at the worksheet footers, or you can open the Graph Properties dialog box for the template graph, and click the Plots tab, and then under Settings for, click Data. Note: Graphs created by templates can be modified like any other graph. For more information, see Creating Graphs in Chapter 4.
Applying Templates
There are three methods for using pages as templates: Using a template from the New Page command. Creates a new page with attributes from the template applied. For more information, see Creating a New Page with Attributes from a Template. Copying a graph page from one notebook section to another. Creates a new page in a section, using the data in the existing worksheet for graphs. For more information, see Copying a Graph Page to use as a Template. Overwriting an existing page. Replaces the existing page. For more information, see Overwriting an Existing Page.
2. Select Graph Page from the New drop-down list. 3. Select the type of graph page you want to open from the Type scroll-down list. 4. Click OK.
1. Make the graph page the active window. 2. From the menus select:
File Templates
1. Make the graph page the active window. 2. From the menus select:
File Templates
3. Click Browse. 4. Select the path and file name of the desired SigmaPlot Notebook or template file. 5. Click Open. 6. Select a template from the Templates list in the Templates dialog box. 7. Click Apply.
Template.jnt is the default source for new pages. It contains both pages with no graphs and pages with graphs.
You can modify existing pages or add your own graphs or graph pages to Template.jnt. Open the file, open the page you want to modify, then save your changes. You can add files by creating new pages or by copying pages from your notebooks to Template.jnt. For more information, see Adding Another Graph to a Page above. You can also create your own template notebook containing your own customized graph pages. Save template notebooks as SigmaPlot Template (.jnt) files, then specify that file to be your Template File.
3. Type the path and file name of the desired template file in the Template File field. 4. Click OK. The notebook becomes the default template source. Note: If a valid default template source file is not specified, a default page is created instead. This page is a letter-sized, white portrait page by default.
The Open dialog box appears. 2. Select Template Notebook from the Files of type drop-down list. 3. Select Template.jnt from the SPW9 folder. 4. Click Open. 5. Open or view the notebook file containing the page you want to add to Template.jnt. 6. Select the page you want to copy. 7. Press Ctrl+C. 8. Select the section of Template.jnt where you want to add the new page. 9. Press Ctrl+V. The page is added to Template.jnt. 10. Save and close Template.jnt. 11. From the menus select:
File New
The New dialog box appears. 12. Under New, select Graph page. The page you copied appears on the list.
Pasting Objects
You can paste Clipboard contents to any open page, report, or into any other Windows application that supports Windows Metafiles or OLE (Object Linking and Embedding). To paste an object to a page, click where you want the object to appear, then press Ctrl+V. You can also click the Paste button on the Standard toolbar, or choose the Edit menu Paste command. For more information, see Using OLE to Paste, Link and Embed Objects below. Note: The Clipboard is a Microsoft Windows feature. To learn more about how the Clipboard works, refer to your Windows Users Guide.
created with the Microsoft Word Equation Editor into a SigmaPlot report, and edit it with the Word Equation Editor when it changes.
Figure 5-11 Example of an Microsoft Excel Equation Embedded into a SigmaPlot Report
Paste Special Command. Allows you to choose Clipboard file types and to also embed objects and links. Insert New Object Command. Directly creates and places an OLE object without using the Clipboard. Allows embedding the object or linking to a file. Drag and Drop. Moves, or copies any Clipboard object (usually OLE).
2. Select the type of object to insert from the Object Type list. A description of the object type appears below. 3. Click OK to insert the object.
View as Icon
With OLE, the View as Icon allows you to place an icon representing the application that created the file in your data. For example, if you have a description of a graph written in a Microsoft Word document, you can embed it, and display it as an icon that shows on the graph page. If you want the object displayed as an icon, select the Display As Icon option. Click the icon to view and edit the object in its source application.
Figure 5-12 Displaying a Microsoft Word Document as an Icon on a Graph Page
The Insert Object dialog box appears. 2. Select the type of object to insert from the Object drop-down list, and click OK. 3. Select Display as Icon. 4. Click OK to insert the object as an icon.
1. Select the graph to cut or copy. 2. Press Ctrl+X or Ctrl+C . The graph is cut or copied. 3. Open or switch to the other application, and click where you want the graph to appear.
4. Paste the graph, typically using the Edit menu Paste command . If the graph is not an OLE object, try the Paste Special command, and select SigmaPlot Graph or SigmaPlot Graph Object. 5. To create a link between SigmaPlot and the other application, click the Paste Link button. To insure you are pasting an OLE object, use the Paste Special command. If a Paste Special command does not exist, the application probably does not support OLE.
Figure 5-13 Using the Paste Special Dialog Box to Paste a Graph from SigmaPlot to another program
The SigmaPlot graph appears in the other application. 6. You can now in-place activate the graph by double-clicking it, or open it in SigmaPlot, by choosing the Edit menu Object command. If the application does not support OLE, the SigmaPlot graph is pasted as a metafile or bitmap graphic. SigmaPlot graphs pasted with the Edit menu Paste command take their plotted data with them in the form of the plotted graph (the worksheet is not shown). If you want to view or edit the data, you must open the graph rather than simply editing it.
1. Open the application and file containing the desired artwork or text, and cut or copy the object. 2. Switch to SigmaPlot and view the graph or report page. 3. Click the location where you want the object to appear, then press Ctrl+V. The graphic is pasted to the page. If the object can be an OLE object, SigmaPlot always defaults to the OLE object. 4. To paste the object as a specified file type, choose the Edit menu Paste Special command. The Paste Special dialog box appears.
Figure 5-15 Using the Paste Special Dialog Box to Paste an Object from Microsoft Word to SigmaPlot
Note: The options available in the Paste Special dialog box depend on the type of file being pasted. 5. If you want the object displayed as an icon, click Display As Icon. Click the icon to view and edit the object in its source application. You can also specify a different icon to display the pasted object. Click Change Icon and select a different icon. 6. Click Paste to paste the object as a specified file type. Select Paste Link to paste the object as a linked file that can be updated in another application. The options in the As list change depending on your selection of either Paste or Paste Link, and the explanation in the Result box changes depending on your selection in the As list. 7. Select the type of object to paste from the As box, then click OK. The object appears at the selected location.
1. View the report or graph page, and click where you want the insertion point. 2. From the menus select:
Insert New Object
The Insert Object dialog box appears. 3. If you want to display the new object as an icon, select Display As Icon. You can also specify a different icon to display the inserted object. Click the Icon button to open the Change Icon dialog box. Choose a different icon from the available options, or click the Browse button to search for alternative icons on your system. 4. To create a new object to place on the report or graph page, select Create New, then choose the type of object from the Object Type list. Click OK to open the application associated with the selected object. Create the desired object, then use the applications appropriate Exit command to close the application and return to SigmaPlot. The created object is displayed on the graph or report page as an embedded object.
Figure 5-16 The Insert Object Dialog Box
5. To insert an object from an existing file on the report or graph page, select Create from File, then type the path and file name of the desired file in the File edit box, or click the
Browse button to open the Browse dialog box, from which you can select the appropriate path and file name of the object you want to place. 6. Select the Link option to place the object on the page as a linked object. When a file is linked, it is modified in your graph or report page when it is modified in the original application. If you did not select theLink option, the object is pasted as an embedded object.
Figure 5-17 The Insert Object Dialog Box After Selecting Create From File, with the Display as Icon Option Checked
7. Click OK.
1. View the graph or report page by selecting it. 2. On the Edit menu, click Links. The Links dialog box appears displaying the path, file name, type of file, and if it is a manually updated or automatically updated link, of all links on the page.
If you do not have any linked objects on the page, the Links box is empty. 3. To change the updating to either Automatic or Manual, select the unselected option. If Automatic updating is selected, the object changes automatically when the source file is changed. If Manual updating is selected, you must click Update Now to update the linked object with any changes made to the source file. 4. To edit a linked object, select the object name in the Links dialog box, then click Open Source. The source file opens in the appropriate application where you can make changes, then exit the application and return to SigmaPlot. If Automatic updating is selected, the object reflects the changes; if Manual updating is selected, you must click the Update Now button to apply changes to the linked object. 5. To change the source file for a linked object, click Change Source. Choose the new path and file name, then click OK. The link appears in the Links dialog box with the new path and file name. You may need to click the Update Now button to view this change in your document.
Figure 5-19
6. To end the link between an object and its source file, click Break Link. The object is no longer treated as a linked object. 7. Click OK to close the Links dialog box.
Figure 5-20
Note: You can also drag and drop graphs onto your Windows desktop. Dropping a graph onto the desktop creates a scrap file that can be dragged into another document at a later date.
1. Open the source and destination pages. The pages must still be within the same notebook, but can be in different sections. 2. Select the graph and drag it from the original page to the new page. If you want to copy rather than move the graph, press the Ctrl key while dragging. 3. Release the mouse where you want the graph to appear. The graph is placed on the new page. If the page is in a different section, the data plotted by the graph is copied to the current worksheet.
2. Click the Page Layout tab. The graphs on the current page are listed in the Sh?wn box. 3. To hide a graph, select it from the list and click Hide. The selected graph is moved to the Hidden list. (To select multiple graphs, hold down the Shift or Ctrl key while making selections.) 4. To view a hidden graph, select it from the Hidden list and click Show. 5. Click OK to apply your selections and close the Page Setup dialog box.
1. Right-click the legend or title and on the shortcut menu, click Hide. The title or legend is not deleted, only hidden. 2. You can also hide graph titles, axis title, and legends using the Graph Properties dialog box. Open the Graph Properties dialog box by double-clicking the graph. You can also right-click the graph, and on the shortcut menu, click Graph Properties. 3. Click the Graph tab. 4. Under Settings for, select Legends. 5. To hide the graph title, clear Show Title. 6. To hide the automatic legend, under Legend properties, clear Show Legend. 7. To hide axis titles, select the Axes tab, under Settings for, click Labels, and clear the Show Axis Title option(s). 8. Click OK to apply the changes and to close the Graph Properties dialog box. The titles and automatic legend no longer appear on the graph page. Restore the title and legend by returning to the Graph Properties dialog box and checking the Show Title and Show Legend options.
Deleting removes curves, plots and graphs entirely. You can use undo (Ctrl+Z) to retrieve these items. When a graph or plot is removed, worksheet data remains intact. Delete also completely removes drawn and pasted objects. Note that delete only hides titles and legends, and does not remove them permanently.
The drawing tools on the Page toolbar buttons are: Select Object. Use the Select Object button to select objects on the graph page. Draw Line. Click this button to draw a line on the graph page. Draw Arrow. Click this button to draw an arrow on the graph page. Draw Box. Use the Draw Box button to draw a box on the graph page. Draw Ellipse. Click this button to draw an ellipse on the graph page. Text. Click this button to add text, labels, or manually created legends to the graph page.
Drawing an Object
To draw an object:
1. Click a drawing tool on the Page toolbar, or choose a drawing command from the Tools menu. 2. The pointer has a crosshair appearance when over the graph page. Place the pointer over the page where you want the object to begin, press and hold down the left mouse button, then drag the pointer to draw the object. 3. Release the mouse button to finish drawing the object.
1. Select the object(s) to modify on the graph page. 2. From the menus select:
Format Fill
3. Click the Fills tab. 4. From the Background Color drop-down list, choose a color. 5. Click OK to apply your changes and to close the dialog box.
Changing Lines
For drawn lines and graph lines, you can change line type, color, and thickness. You can also use the Object Properties dialog box to add arrowheads and other line endings to lines. For more information, see Using Custom Colors below.
To change line color:
3. Under Line, select a color from the Color drop-down list. Choose None to create a transparent line. 4. Click OK to apply your changes and to close the dialog box.
To change line type and thickness:
1. Select the object(s) to modify: 2. On the Format menu, click Line. The Object Properties dialog box appears. 3. Click the Line tab.
4. To set the line type, under Line, select a type from the Type drop-down list. 5. To set the line thickness, use the Thickness slider. Clicking the slider causes the slider to move incrementally, while dragging it moves it more precisely. To change the range of control of the slider, move the slider to one end of the selectable range, select the text in the corresponding edit box, and type a new numeric value. 6. Click OK to apply your changes and to close the dialog box.
Figure 5-25 The Line End tab of the Object Properties Dialog Box
3. Click the Line End tab. 4. Add or edit line ends at both the start and end of a line. The Start options add or modify the beginning end of the line (where you start drawing the line). The End options add or modify the line end at the end of the drawn line (where you stop drawing the line by releasing the mouse button). 5. To change the type of line used, select a style from the Style drop-down list. 6. To change the arrowhead length and angle, move the Angle and Length slider. The angle is the angle between the arrowhead line and the main line. The Angle option is unavailable if the line Style is dotted or plain. Note: Clicking the slider causes the slider to move incrementally, while dragging it moves it more precisely. To change the range of control of the slider, move the slider to one end of the selectable range, select the text in the corresponding edit box, and type a new numeric value. 7. Click OK to apply your changes and to close the dialog box.
2. Drag it to the desired position. A dotted outline of the graph follows the pointer indicating the location of the moved graph. 3. Release the mouse button. The graph moves to the new position.
1. View the page window. 2. Click the graph or desired objects to select them. Selected page objects are surrounded with small square handles. Place the pointer over a handle. 3. Press and hold down the left mouse button to drag the handle to a new location. The shape of the pointer changes when you move it over a handle, indicating the direction you can stretch the graph or object. Drag a side handle to stretch or shrink an object horizontally, drag a top or bottom handle to stretch or shrink an object vertically, or drag a corner handle to stretch an object two-dimensionally. A dotted outline of the resized graph or object follows the pointer position.
Dragging a corner handle preserves the aspect ratio (relative height and width) of objects by default. Also, graph text, symbols and tick marks are rescaled along with the graph. To disable these features, use the Tools menu Options command and change these Page option settings. For more information, see Setting Page Options above. 4. Release the mouse button when finished. The graph or object resizes to the indicated size. Note: Unlike graphs and drawn objects, you cannot stretch or shrink text labels manually. To resize text, change the font size. For more information, see Formatting Text below.
1. Select the graph or object on the page by clicking it. 2. Right-click the selected item, and on the shortcut menu, click Object Properties. The Object Properties dialog box appears.
Figure 5-28 Object Properties Dialog Box Size and Position Tab
3. Click the Size and Position tab. 4. To set the distance of the selected object from the top and the left of the page, under Position, move the Top and Left sliders or type new values in the Top and Left boxes. 5. To change the size of the selected object, under Size, move the Height and Width sliders to set the size to specific measurements, or scale the object to a new size by typing a percentage in the Height and Width boxes. 6. Click OK.
Pressing an arrow key moves the graph or object one one point, or .014in. You can change this default setting in the spw.ini file. If you have activated Snap-to grids, nudge will not work unless you set the nudge value to be greater than or equal to the Snap-to value. You cannot nudge computable objects, such as plots and all parts of plots, tick marks, and regression, reference, and grid lines.
1. Select the object to move by clicking it. 2. To move the selected object to the foreground, on the Page toolbar, click Bring to Front. The selected object is drawn in front of all other objects. 3. To move the selected object to the background, on the Page toolbar, click Send to Back. The selected object is drawn behind all other objects. Note: If you select more than one object, the selected objects remain in their relative front to back positions. Grouped objects, including titles and legends with graphs, move as a single object.
2. Select the graph, by clicking it, if you wish to attach the graph to the objects or text. 3. Select the objects and text to group by holding down the Shift key while selecting individual objects. Handles appear around the graph and each selected object. 4. On the Page toolbar, click Group. The Group command and button are available only when more than one object is selected. All selected objects are grouped and can be selected, moved, sized, aligned, and positioned as a single object.
To ungroup objects on a graph page:
If you have grouped a group, you may need to ungroup the objects as many times as they have been grouped.
3. From the Layouts list, select a layout for the page. A preview of the layout appears in the Preview window. Layouts are stored in a template file called Layouts.jnt. For more information, see Adding New Pages to Layout.jnt below. Note: You must apply a layout to a page that has the same or fewer number of graphs. 4. Click Apply. The graphs on the page match the layout you selected, and the Layout dialog box remains open. 5. To arrange the graphs again, you can select another layout from the Layouts list, then click Apply, or click Close to close the dialog box.
2. Select Template Notebook (*.jnt) from the Files of type drop-down list.
3. Select Layout.jnt from the SPW_10 folder. For more information, see About SigmaPlots User and Program Files in Chapter 1. 4. Click Open. The Layout.jnt notebook appears in the Notebook Manager.
Figure 5-31 Layout Notebook
5. Open or view the notebook file containing the page you want to add to Layout.jnt. 6. Select the page you want to copy. 7. Press Ctrl+C. 8. Select the section of Layout.jnt where you want to add the new page. 9. Press Ctrl+V. The page appears in Layout.jnt and also at the bottom of the section. 10. On the File menu, click Save to save the notebook.
3. Type the name of the new layout template notebook in File name box. 4. Select SigmaPlot Template File from the Save as type drop-down list. 5. Click Save. Now you can add future layouts to their own separate layout notebook.
2. Click the General tab. 3. Type the path and file name of the desired layout file in the Layout file field. 4. Click OK. The notebook becomes the default layout source.
1. Select the labels, graphs or other object(s) you want to align by holding down the Shift key while selecting individual objects. (You must select more than one object to use the Align command.) 2. From the menus select:
Page Align
The Align dialog box appears. 3. Under Horizontal and Vertical, choose the appropriate options to align the selected objects vertically, horizontally, or both. Graphical feedback for your selections appears in the lower right corner of the dialog box. 4. To align selected objects relative to each other, select Each Other. You must have multiple objects selected if you want to align selected objects relative to each other. Each Other moves aligned objects with respect to the last selected object, which remains in a fixed position. The last selected object can be distinguished from other selected objects by solid rather than hollow selection handles. 5. To align objects relative to the page margins rather than the page edge, select Page Margins. Note: If you select Page Margins, objects will not move with respect to each other. You can select Page Margins to place single objects. To set margins for each page, on the File menu, click Page Setup. 6. Click OK.
Using Rulers
Rulers are optionally displayed at the top and left hand side of all graph pages. They display the current units set in the Tools menu Options dialog box. You can choose between inches, centimeters, or points.
Using Snap-to
You can use Snap-to if the grids are displayed or hidden. Select Snap-to in the Tools menu Options dialog box, or right-click the graph page and on the shortcut menu, click Snap-to. Graphs and objects snap to the nearest grid.
Using Crosshairs
Use Crosshairs as an object alignment tool. To turn on crosshairs, click the Crosshairs button on the upper left hand corner of the graph page window. Crosshair lines extend from the pointer tip to the rulers and to the right and bottom of the window, and follow the pointer. To hide crosshairs, click the Crosshairs button again.
Editing Text
Use the Page toolbar to add and edit text labels and legends to the graph page, in addition to editing automatically created graph and axis titles. SigmaPlot automatically creates legends for every plot. You can modify the existing automatic legend by clicking the Text button on the Page toolbar, and then edit the text using the Formatting toolbar. You can format tick and contour labels, but you cannot edit their content.
Mixed fonts and other attributes within a single label. Multiple levels of superscripting and subscripting. Rotation of text in single degree increments. Color using up to 16.7 million different combinations of red, green, and blue
To create text labels or legends on a page:
1. Select and view the page window, then click the Text button on the Page toolbar. This places you into text mode until another mode or tool is selected 2. Click the page where you want the label to begin. A text box appears. 3. Select the font, character size, and other starting character attributes from the Formatting toolbar. The Rotation, Alignment, and Line Spacing options affect the entire label, not just the selected text, and Line Spacing is a minimum spacing control, not fixed. If you change the height of characters by changing font sizes or by adding superscripts or subscripts, the line height adjusts automatically. Note: Using the Default Text Properties you can set default text label attributes by opening the Text Properties dialog box with no labels selected. For more information, see Formatting Text below. Note: In addition to using the Greek Characters button to add a Greek symbol to text, you can also select pre-existing text and choose Symbol as the font type in the Text Properties dialog box. For more information, see Formatting Text below. 4. Type your label. 5. To type additional lines, insert a line break by pressing Enter. 6. To change the attributes of text already typed in the Edit Text dialog box, drag the cursor over the text you want to change to highlight it, then click the appropriate button, such as normal font, bold, italics, underline, sub or superscript, or symbol. 7. To switch back to normal text from greek, superscript, or subscript text, click the Normal button.
8. To add legend symbols to your text, click Symbols. The Symbol palette appears.
Formatting Text
If you want only to change the attributes (the formatting) of selected text on a graph page, use the Formatting toolbar. The Text Properties dialog box sets properties for all selected labels, and applies changes to all characters within selected labels. Note: If you have complex font and character changes within a label, take care not to overwrite these formats with Text Properties dialog box settings. Global Text Changes. The Text Properties dialog box is useful for formatting multiple labels as well as all text on a graph. Select the graph and choose Text Properties, then select the attributes you want applied to all graph labels and titles. Default Text Properties. The Text Properties dialog box is used to set the default character and paragraph properties for new labels. Open the Text Properties dialog box with nothing selected, and set the options you want applied to new text labels.
To format text using the Text Properties dialog box:
1. Select the text object you want to modify. If you want to modify several text objects, hold down the Shift key while clicking the objects, or drag a select window around all objects 2. From the menus select:
Format Text Properties
Figure 5-34
3. To change the font, style, character size, or color of text, or to underline text, click the Font tab. Note: If you have multiple text objects with different text properties selected, the attributes that are not the same appear blank. Do not select an attribute for these options unless you want it to be applied to all selected objects. 4. To change paragraph attributes, including line spacing, alignment, or rotation, click the Paragraph tab.
294 Chapter 5
1. View the page. On the Page menu, click the Text button. 2. Double-click the legend entry that you want to edit. 3. Edit the text of the legend entry as desired using the Formatting toolbar. You can also change the legend symbol properties, including Symbol size, by clicking the Symbol button. For more information, see Sizing Legend Symbols below.
You cannot change the widths of automatic legendsthese are determined automatically by the width of the text. You can edit individual labels and add multiple lines. You can also ungroup a legend and format it manually. For more information, see Ungrouping a Legend below.
1. Double-click the legend. 2. On the Formatting toolbar, click the Symbol button. The Symbol dialog box appears.
296 Chapter 5
Figure 5-38 Using the Symbol Dialog Box to Resize Legend Symbols
3. Under Symbol, select the symbol to use for the label. This list displays all symbols, lines and fills used by the selected graph source. 4. Under Size, move the Width and Height sliders to increase symbol size, or enter a symbol size value. The Width value determines the space between symbols, while the Height value determines the actual symbol size. This means the larger the height, the larger the symbol size; the larger the width, the larger the space between the symbol and text. For line and scatter plots, the width can never be less than the height. 5. Click OK to close the dialog box and save the changes.
1. Double-click the graph to open the Graph Properties dialog box. 2. Click the Graph tab.
Figure 5-39 Use the Graph tab of the Graph Properties dialog dox to Edit an Automatic Legend
3. To show or hide an automatic legend, under Legend Properties, select or clear Show Legend. For more information, see Hiding and Viewing Graphs on a Page above. 4. To enclose the legend in a box, under Legend properties, select Framed in box. 5. To hide a legend box, under Legend properties, clear Framed in box. 6. To modify the line thickness and fill of the legend box, under Legend Properties, click Box to open the Object Properties dialog box. 7. To halt all automatic updating of the legend text for the whole legend, select Lock legend. For more information, see Locking Legend Text below. 8. To show or hide individual legend entries for a specific plot or curve, under Legend appearance, from the For legend symbol list, select or clear a legend entry. 9. To annotate from the For legend symbol drop-down list, enter the text for a legend symbol by selecting the symbol then select the Legend text box and type text. Do this for as many legend symbols as you want.
298 Chapter 5
10. To move the legend symbols either to the right or to the left of text, select a position from the Symbol placement drop-down list. If you have no legend symbol selected, this operates on all legends. If you select a specific entry from the For legend symbol list, this option affects only that symbol. 11. To modify the appearance of the symbols for the current legend, select a symbol style from the Style drop-down list. The Style drop-down list only affects scatter and line plots. If you have no legend symbol selected, this operates on all symbols. If you select a specific entry from the For symbol list, this option affects only that symbol. 12. To change the text size or style, under Legend properties, click Font. The Text Properties dialog box appears. For more information, see Formatting Text above. 13. To restore all legend text and symbols to the default settings, under Legend properties, click Reset. Note: The Reset button also unlocks the legend, if locked. When you click Reset defaults, a Novice prompt appears which you can disable. 14. Click OK to apply the changes and close the Graph Properties dialog box. The legend is updated as specified.
The Options dialog box appears. 2. Click the General tab. 3. Select Automatic legends to display the legend, or clear it to hide the legend.
Ungrouping a Legend
You can ungroup the legend entries and box by selecting the legend, then choosing the Format menu Ungroup command, or clicking the Page toolbar Ungroup button. You can then edit each object like an ordinary graphic object or label. You can also use your mouse to move any of the legend items to a new location,and the Format menu Align command. Note: Ungrouping a legend removes automatic legend features.
300 Chapter 5
Figure 5-40 Use the Graph tab on the Graph Properties dialog box to lock legend text.
2. Click the Graph tab. 3. Under Settings for, click Legends. 4. Under Legend properties, click Lock legend. 5. Click OK to close the dialog box.
Figure 5-41 The Margins Tab of the Page Setup Dialog Box
Note: The options in the Page Setup dialog box affect both the view of the page onscreen, and the printer settings for the page you are printing. For more information, see Printing Graph Pages above.
The Page Setup dialog box appears. 2. Click the Margins tab. 3. Use the Top, Bottom, Left, and Right options to specify the width or height of the corresponding page margin. You can type values in the edit boxes using any of the available units of measurement; the value is converted to the current measurement units specified in the Options dialog box. Type in for inches, mm for millimeters, and pts for points.
302 Chapter 5
Margins do not affect printing, they are only a guide. The Align dialog box uses margins when aligning the page. 4. Clear or check the Show Margins option by selecting it. If this option is checked, margins are displayed on the page. To hide page margins, clear Show Margins. 5. Click OK. For more information, see Changing Page Units of Measurement below.
The Page Setup dialog box appears. 2. Click the Page Size tab. 3. From the Paper Size drop-down list choose the appropriate size for the page, or select unique page sizes from the Width and Height drop-down lists. 4. Note: SigmaPlot does not support heights or widths greater than 32 inches. 5. To switch between portrait (normal) and landscape (sideways) orientation, select either the Portrait or Landscape option. 6. Click OK to accept your changes and close the dialog box. Note: If you change the page size and/or orientation, the page changes on the screen, but your graphs remain in the same relative position. You may have to move the graphs back into position.
304 Chapter 5
2. Click the Page tab. 3. From the Units box, select the unit of measurement to use on the page. You can choose to use inches, millimeters, or points. 4. Click OK to accept the changes and close the dialog box.
1. Make the page active by selecting it, or by choosing its name from the Window menu. A check mark next to the name of the page indicates that the page is active. 2. From the menus select:
File Page Setup
3. Click the Graph Layout tab. 4. From the Color drop-down list, select the color to use for the page. Select (Custom) to use or create a custom color. For more information, see Using Custom Colors below. 5. Click OK. Note: If you want no background color to show up for pasted graphs (e.g., pasting a graph into PowerPoint), set the page color to None.
Templates
You can overwrite the current page entirely by applying a template to it. This is not recommended as a means of reformatting the page unless you intend to discard all changes made to the page up to this point. For more information, see Using Graph Pages as Templates above.
306 Chapter 5
systems support Hi Color (16-bit) or True Color (24-bit) modes. Right-click your desktop, choose Properties, select Settings, then set your Color palette to the highest possible level.
To select a custom color:
1. Open the dialog box that has the color option in it, and from the Color drop-down list, select (Custom).
Figure 5-44 Selecting the Custom Color option from the Text Properties Dialog Box.
You have not already selected a custom color, the Color dialog box appears.
If a custom color has already been defined for this option, the custom color is selected. 2. From the Basic Colors list, select a color, or click Define Custom Colors to define your own color. The dialog box expands to show a color palette. 3. Click the large color field, or drag your mouse across it to indicate the approximate color you want to use. If you know the numeric RGB (red, green, blue) values of the desired color, you can select each of the Red, Green, and Blue edit boxes and type the correct values. The selected color box appears. 4. Move the slider next to the vertical color bar along the right of the dialog box to finetune the range of the Hue, Saturation, and Luminosity of the selected color, or type new values in the edit boxes. The current custom color appears in the Color|Solid box as a gradational color and a solid. 5. To change the color assigned to a Custom Color box, select the box in the list, then specify the new color from the large color field. 6. To select the gradational color, click Add to Custom Colors. The color appears in the first available box of the Custom Colors list.
308 Chapter 5
7. To select the solid version of the color, double-click the solid in the Color|Solid box, then click Add to Custom Colors. The color appears in the first available box of the Custom Colors list. 8. Select the color to use from the Custom Color list, then click OK. The Color dialog box closes, and you are returned to the dialog box from which you opened the Color dialog box. The color drop-down list that you are using now has the color you created as an option with the word (Custom) next to it. If the custom color you created is a duplicate of a pre-existing system color, the system color is selected instead of the (Custom) option in the drop-down list.
Chapter
You can create 2D Cartesian (XY) plots from many worksheet columns or column pairs. Each column is represented as a separate curve, set of bars, or box, depending on the plot type. 2D graphs must have at least one plot, but you also can display many more plots, each with a different type and style. You can draw linear or polynomial regressions with confidence and prediction intervals. For more information, see Plotting and Modifying Regression Lines in Chapter 4. You can also draw reference lines for each curve. For more information, see Adding Reference Lines in Chapter 4. This chapter covers: Creating 2D plots. For more information, see Creating 2D Plots below. Creating 2D scatter plots with error bars. For more information, see Creating 2D Scatter Plots with Error Bars below. Creating 2D scatter plots with asymmetric error bars. For more information, see Creating 2D Plots with Asymmetric Error Bars below. Modifying error bars. For more information, see Modifying Error Bars below. Creating grouped bar charts. For more information, see Grouped Bar Charts below. Creating box plots. For more information, see Creating Box Plots below. Creating area plots. For more information, see Creating Area Plots below. Creating bubble plots. For more information, see Bubble Plots below.
309
310 Chapter 6
Area Plots
Using area plots, you can fill an area under a curve with a color making the curve easier to see. You can orient the fill up, down, left, or right. If your curve is a closed polygon, you can also fill the polygon. You can have multiple curves (plots) on a page, so you can stack Area Plots. For more information, see Creating Area Plots below.
Figure 6-1 Examples of Area Plots
Bar Charts
Bar charts plot data either as vertical or horizontal bars. They originate from zero in either a positive or negative direction. Simple bar charts plot each row of data as a separate bar, and grouped bar charts plot multiple columns of data by grouping data in the same rows. Stacked bar charts plot data as segments of a bar; each data point is drawn as a bar segment starting where the previous data point ended.
Use the Graph Properties dialog box to modify bar width, bar fill colors, and bar fill patterns and to add error bars to simple and grouped bar charts. For more information, see Creating Grouped Bar Charts below.
Box Plots
Box plots graph data as a box representing statistical values. The boundary of the box closest to zero indicates the 25th percentile, a line within the box marks the median, and the boundary of the box farthest from zero indicates the 75th percentile. Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles. In addition, you can graph the mean and outlying points. For more information, see Creating Box Plots below. You need a minimum number of data points to compute each set of percentiles. At least three points are required to compute the 25th and 75th percentiles, and at least nine points are required for the 5th, 10th, 90th and 95th percentiles. If SigmaPlot is unable to compute a percentile point, the related graph element is not drawn.
Creating 2D Plots
To create a 2D plot: 1. Select the worksheet columns to plot before creating your graph by dragging the pointer over your data. 2. Select the desired graph type and style from the Graph toolbar. The Graph Wizard appears.
312 Chapter 6
Figure 6-2 Using the Graph Wizard to Specify the Data Format
3. From the Data Format list, choose the appropriate data format, and click Next. 4. Specify which worksheet columns correspond to the data for your plot. Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in the dialog box and you can click Finish to create the graph. Note: If you have not already picked columns, note that a single data type is highlighted in the Selected Columns list. This shows the data type you are picking a column for. Begin picking data either by clicking the corresponding column directly in the worksheet, or choosing the appropriate column from the Data Columns list. Repeat this process for every column you are using to create your graph. 5. If you make a mistake while picking the data, select the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the list. 6. When you have finished picking data, click Finish to create the plot and close the Graph Wizard. Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph. For more information, see Creating Graphs in Chapter 4.
314 Chapter 6
Figure 6-4 The data in this worksheet is arranged as category (or indexed) data. The data is organized rowwise by categories
1. On the Graph menu, click Create Graph. The Type panel of the Graph Wizard appears. 2. Select one of the following graph types: Line Plot Scatter Plot Line and Scatter Plot 3. Click Next. The Style panel of the Graph Wizard appears. If you selected Line Plot as the graph type, then select any one of the following graph styles: Multiple Scatter Multiple Regression Multiple Spline Curves Multiple Vertical Step Plot Multiple Horizontal Step Plot Multiple Vertical Midpoint Step Plot
Multiple Horizontal Midpoint Step Plot If you select either Scatter Plot or Scatter and Line Plot as the Graph Type, then select any one of these graph styles: Multiple Straight Lines Multiple Spline Curves Multiple Vertical Step Plot Multiple Horizontal Step Plot Multiple Vertical Midpoint Step Plot Multiple Horizontal Midpoint Step Plot 4. Click Next. The Data Format panel of the Graph Wizard appears.
Figure 6-5 Selecting a Category data format from the Graph Wizard
5. Select one of the following category data formats: XY Category. Uses one worksheet column to graph the categories, and a pair of XY columns. X Category. Uses one X column, and a column for categories, indexes, or levels to group the data in corresponding rows. Y Category. Uses one Y column, and a column for categories, indexes, or levels to group the data in corresponding rows. 6. Click Next. The Select Data panel of the Graph Wizard appears.
316 Chapter 6
7. Select the data columns which corresponds to its axis or category. For example, if you are using an XY Category Data format, first select the column to use for the X data from the drop-down list. This selection appears in the Selected columns list. Then select the column to use for the Y data from the drop-down list. Lastly, select the column to use as Categories from the drop-down list.
Figure 6-6 Selecting columns of data for category or grouped data
8. If you make a mistake while picking data, select another entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the list. 9. When you have finished picking data, click Finish to create the plot and close the Graph Wizard.
Figure 6-7 Legends for category data plots include the title of the two columns containing the observation data and also text describing which category group each symbol pertains to
Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph. For more information, see Creating Graphs in Chapter 4.
318 Chapter 6
To add error bars to an existing plot, change the plot type. For more information, see Changing Graph Type and Style in Chapter 4.
To create a scatter plot with error bars:
1. Select the worksheet columns to plot before creating your graph by dragging the pointer over your data. 2. On the 2D Graph toolbar, click Scatter Plot, and then click Simple Scatter Error Bars. The Graph Wizard appears. 3. From the Symbol Value drop-down list, select the error bar source. Symbol Value. Choose either Column Means to use the column means as the error bar source, Replicate Row Means to use the row means as the error bar source, Worksheet Columns to use the values you have entered in the worksheet, or Worksheet Columns to read error bar end values from sets of two adjacent columns. You are prompted during data picking to specify the column to use as error bar source data.
Error Calculation. If you choose any option besides Worksheet Columns as the symbol value, specify the error calculation method to use for upper and lower error bars.
Figure 6-10 Selecting the Error Bar Source
4. From the Error Calculation - Upper and Error Calculation - Lower drop-down lists, specify the error calculation for the error bars. Error Calculations are not applicable if you select Worksheet Columns or Asymmetrical Error Bars from the Symbol Value list. 5. Click Next.
320 Chapter 6
6. From the Data Format list, select the appropriate data format. X column averaged plots require a constant Y column value, and Y column averaged plots require a constant X column value. 7. Click Next.
Figure 6-12 Specifying the Data Columns for the Error Bars
8. Specify which worksheet columns correspond to the data for your plot. Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in the dialog box, and you can click Finish to create the graph.
9. To create a single plot graph, choose data for every column you are using to make the graph. To create a graph of multiple plots, choose data for the first plot, then click Next to pick the data for the next plot. Repeat this process for as many plots as necessary. 10. To make a graph with simple error bars or a graph with multiple error bars using worksheet columns as the Symbol Value for error bar data, you are prompted to choose columns for error bar data. Repeat the data picking process for every column you are using to create your plot. For more information, see Computing Percentile Methods below. 11. To make a graph using any of the other sources for error bar data (i.e. Column Means, Column Median, Standard Error, etc.) with multiple error bars, you can create a graph using a single plot, or a graph with multiple plots. Use multiple plots if you want to use different symbols to distinguish between data sets. Note: If you make a mistake while picking data, click the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list. Click Back to access previous Graph Wizard panels. 12. Click Finish when you have finished picking the data to create the plot.
1. Select the worksheet columns to plot before creating your graph by dragging the pointer over your data. 2. On the 2D Graph toolbar click Scatter Plot and then Simple Scatter - Error Bars. The Graph Wizard - Create Graph dialog box appears. 3. Select Column Means or Column Median from the Symbol Value drop-down list. 4. Select Maximum from the Error Calculation - Upper drop-down list.
322 Chapter 6
5. Select Minimum from the Error Calculation - Lower drop-down list. 6. Click Next. The Graph Wizard prompts you to select a data format. 7. Select X Many Y from the Data Format list, and click Next. Since you have already selected the data columns to plot, the appropriate column titles appear in the Selected Columns list. 8. Click Finish. A range plot appears.
1. Drag the pointer over your worksheet data to select the data. 2. On the 2D Graph toolbar, click Scatter Plot, and then click either Simple Scatter Vertical Asymmetrical Error Bars or Simple Scatter - Horizontal Asymmetrical Error Bars.
The Graph Wizard appears. 3. From the Data Format list, select a data format, and click Next. 4. Specify which worksheet columns correspond to the data for your plot. For more information, see Picking Different Data for the Current Plot in Chapter 4. 5. Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. 6. Click Finish to create the graph. For more information, see Modifying Error Bars below.
324 Chapter 6
1. Select the worksheet columns to plot before creating your graph by dragging the pointer over your data. 2. On the 2D Graph toolbar click Scatter Plot and then Multiple Scatter - Error Bars. The Graph Wizard - Create Graph dialog box appears. 3. Select Column Median from the Symbol Value drop-down list. 4. Select 75th Percentile from the Error Calculation - Upper drop-down list. 5. Select 25th Percentile from the Error Calculation - Lower drop-down list. 6. Click Next. The Graph Wizard prompts you to select a data format. 7. Select X Many Y from the Data Format list, and click Next. Since you have already selected the data columns to plot, the appropriate column titles appear in the Selected Columns list. 8. Click Finish.
1. Open or import a worksheet using a category data format. For more information, see Importing Files from Other Applications in Chapter 3. 2. On the Graph menu, click Create Graph. The Graph Wizard - Create Graph - Type dialog box appears.
3. Select a graph type from Graph types list, and click Next. The Graph Wizard - Create Graph - Style dialog box appears. 4. Select a graph style that uses error bars from the Graph styles list, and click Next. The Create Graph - Error BarsGraph Wizard - dialog box appears. 5. Select either Category Mean or Category Median from the drop-down list. 6. Select error calculations from the Error calculation - upper and Error calculation lower drop-down lists, and click Next. The Graph Wizard - Create Graph - Data Format dialog box appears. 7. From the Data for Categories drop-down list, select a column that corresponds to the categorical data you wish to plot. 8. From the Data for Y drop-down list, select the column that corresponds to the Y data you wish to plot, and click Finish. An error bar plot appears.
326 Chapter 6
1. Double-click the plot. The Graph Properties dialog box appears. 2. From the Settings for list, select Error Bars.
Figure 6-16 Graph Properties Dialog Plots Tab Error Bar Setting
3. To change the color of the error bars, from the Line Color list, select a line color. 4. To change line thickness and error bar cap width, move the Thickness and Cap Width sliders.
5. Click OK.
2. Click the Plots tab. 3. From the Plot drop-down list, select the plot with error bars to modify. 4. From the Settings for drop-down list, select Error Bars.
328 Chapter 6
5. Under Error Bars, from the Direction drop-down list, select the direction of Y. 6. Select either X or Y Positive or Negative. Note: An X positive absolute direction always points right; a Y positive direction always point up. An X negative absolute direction always points left; a Y negative absolute direction always points down.
Figure 6-18 The bar chart on the left uses Y error bars with an absolute positive direction. The bar chart on the right uses a relative direction away from zero.
7. Double-click the plot. The Graph Properties dialog box appears. 8. Click the Plots tab. 9. From the Settings for list, select Error Bars. 10. From the Error Bars Direction drop-down list, select X or Y From Zero or To Zero. Note: A relative to zero direction always points toward or away from zero. This option is useful for bar charts that have negative values.
Figure 6-19 The bar chart on the left uses X error bars with an absolute negative direction. The bar chart on the right uses a relative direction towards zero.
330 Chapter 6
2. Enter the codes for the error bar directions. The codes for the directions are
Table 2-1DirectionCode
Absolute Positive
Positive or P
Absolute Negative Relative From Zero Relative To Zero Absolute or Relative, Both Directions
Note: Codes you type in the worksheet can be either upper or lower case. 3. Double-click the plot. The Graph Properties dialog box appears.
Figure 6-21 Setting Error Bars
5. From the Settings for list, select Error Bars. 6. Under Error bars, from the Direction list, choose the name of the first column which contains the error bar direction codes. Note: SigmaPlot assumes that it is the next column that contains the second column of error bar codes. 7. Click OK to apply the changes and close the dialog box.
1. Double-click the plot with a log axis scale to open the Graph Properties dialog box. 2. Click the Plots tab. 3. From the Settings for list, select Error Bars. 4. From the Mean Computation drop-down list, select Arithmetic or Geometric.
332 Chapter 6
Figure 6-22 Selecting Arithmetic or Geometric from the Mean Computation list
5. Click OK.
1. Select the plot to modify by clicking it. Small, square, black handles surround the selected plot.
2. On the Standard toolbar, click the Graph Wizard button. The Graph Wizard appears.
Figure 6-23 Graph Wizard - Modify Plot
3. Click Next. 4. From the Data for Error drop-down list, select a column as a new error bar source.
Figure 6-24 Choosing the New Error Bar Source from the Data for Error drop-down list.
5. Click Finish. The graph appears with the new error bars.
334 Chapter 6
The order of the column pairs in the list determines the order of the bars for each group. To change the bar orders within groups, change the order the column pairs appear in the list by using the Graph Wizard to repick column data. For more information, see Picking Different Data for the Current Plot in Chapter 4. Use the Graph Wizard to create grouped bar charts with or without error bars. If creating a grouped bar chart with error bars, error bar values must be from worksheet column values entered prior to creating the plot. You are prompted during graph creation for error bar worksheet columns.
Figure 6-25 Examples of Grouped Bar Charts with and without Error Bars
1. Select the worksheet columns to plot before creating your graph by dragging the pointer over your data. For more information, see Picking Different Data for the Current Plot in Chapter 4. 2. On the 2D Graph toolbar, click Horizontal or Vertical Bar Chart, and then click either Grouped Bar Chart, or Grouped Error Bars. The Graph Wizard appears.
Figure 6-26 Using the Graph Wizard to Specify the Data Format
3. From the Data Format list, choose the appropriate data format to specify how your data is formatted. The data formats available depend on the graph type and style. 4. Click Next. Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. To change the selected data, select the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list.
Figure 6-27 Using the Create Graph Dialog to Pick Columns to Plot.
5. Click Finish. For more information, see Modifying Error Bars above.
336 Chapter 6
2. Select the block of data in the statistics window that consists of the means and standard deviations of the first set of bars. 3. Right-click, and on the shortcut menu click Copy. 4. Select the first row of an empty column in the worksheet. 5. From the menus select:
Edit Transpose Paste
The first pasted column of data is the mean, and the next column is the standard deviations. For more information, see Switching Rows to Columns in Chapter 3.
338 Chapter 6
Figure 6-30 The data in columns 13 and 14 of the worksheet are transposed from the selected data in rows 1 and 2 of the Column Statistics window. Column 13 contains the means of the column data and column 14 contains the standard deviations of the data.
6. Repeat the copy and transpose paste procedure for the remaining sets of bars. Each pair of mean and standard deviation columns you create adds an additional bar to each group. 7. To plot the results, on the 2D Graph toolbar, select a vertical or horizontal bar chart graph type with grouped error bars, then select the desired data format. If you already have a graph, repick the plotted data by selecting the plot to modify, then clicking the toolbar button. 8. If you select X Many Y as the data format, pick your constant value column (either a row number or a single column), then pick the first column of means as your data column, and the first column of standard deviations as the associated error bar column. 9. Continue picking the mean columns and error bars for each set. 10. Click Finish.
Figure 6-31 Picking Data to Plot for a Grouped Bar Chart with Error Bars
1. Select the worksheet columns to plot by dragging the pointer over your data. 2. On the 2D Graph toolbar, click Box Plot and then click Horizontal Box or Vertical Box. The Graph Wizard appears.
340 Chapter 6
3. From the Data Format list, choose the appropriate data format, and click Next.
Figure 6-33 Graph Wizard - Select Data
Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. Note: You need a minimum number of data points to compute each set of percentiles. At least three points are required to compute the 25th and 75th percentiles, and at least nine points are required for the 5th, 10th, 90th and 95th percentiles. If SigmaPlot is unable to compute a percentile point, that set of points is not drawn. 4. Click Finish to create the graph.
Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph.
2. Click the Plots tab. 3. From the Settings for list, select Box Options. 4. To display a mean line in addition to the median line, under Box Plot Mean Line, select Display Mean Line. If the check box is clear, the mean line is not displayed.
342 Chapter 6
5. To modify the mean line, under Box Plot Mean line, from the Line Type drop-down list, select a mean line type. 6. Select a line thickness and color using the Thickness and Color options. Selecting (none) from the Line Type or Color lists creates a transparent mean line. Selecting (Custom) from the color list enables you to use a custom mean line color, or to create a new color. 7. To change how outliers are handled, from the Handling Outliers drop-down list, select either Show Each Outlier (to plot outside the 10th and 90th percentiles), or Show 5th/10th Percentiles (to plot only the 5th and 95th percentiles as symbols). Note: At least nine data points are required to compute the 5th, 10th, 90th and 95th percentiles. Also, there may be no data points beyond the 10th and 90th percentiles. 8. To modify whisker cap width, under Whisker Caps, move the Width slider, or type a new value in the Width box. 9. Click OK.
The Options dialog box appears. 2. Click the General tab. 3. From the Percentile Method drop-down list, select either: Cleveland Standard Both the Cleveland method and the Standard method use linear interpolation to determine the percentile value, but each uses a different method of rounding when determining the smallest data index used for the interpolation. The two methods give the same result when computing the 50th percentile (median). If the data in increasing order is x1, x2, ..., xN and the percentile is p, then the two methods compute the data percentile value v using the following formulas: Cleveland: Let k be the nearest integer to N*p/100, and let f = N*p/100 + .5 - k. Standard: Let k be the largest integer less than or equal to (N+1)*p/100, and let f = (N+1)*p/100 - k. 4. To compute the percentile value, each of the above methods uses the formula: v=f*xk+1+(1-f)*xk. 5. Click OK.
344 Chapter 6
Figure 6-35 This example is actually four plots: a simple straight line, simple scatter, vertical area, and multiple area. You can find this example in Samples.jnb.
Figure 6-36 In this example, there are see two vertical area plots, a simple area plot, and a simple scatter plot.
1. Select the worksheet columns to plot by dragging the pointer over your data. 2. On the 2D Graph toolbar, click Area Plot and then click Simple Area Plot. The Graph Wizard appears.
Figure 6-37 Graph Wizard - Data Format
3. From the Data Format list, choose the appropriate data format, and click Next.
346 Chapter 6
Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. Note: You can plot no more than 2500 data points per curve. 4. Click Finish to create the graph.
Figure 6-39 Example of a Vertical Area Plot
Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph.
1. Select the worksheet columns to plot by dragging the pointer over your data. 2. On the 2D Graph toolbar, click Area Plot, and then click Multiple Area plot. The Graph Wizard appears.
Figure 6-40 Graph Wizard - Data Format
3. From the Data Format list, choose the appropriate data format, and click Next. Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. To change the selected data, select the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list.
348 Chapter 6
Note: You can plot no more than 2500 data points per curve. 4. Click Finish to create the graph.
Figure 6-42 Example of a Multiple Area Plot using a Y Many X data format.
Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph. You can identify intersections either by using the Graph Properties dialog box or by creating a complex area plot. For more information, see Creating Complex Area Plots below.
1. Select the worksheet columns to plot by dragging the pointer over your data. 2. On the 2D Graph toolbar, click Area Plot, and then click Complex Area Plot. The Graph Wizard appears.
Figure 6-43 Graph Wizard - Data Format
3. From the Data Format list, choose the appropriate data format, and click Next. Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. To change the selected data, select the wrong entry in the Graph Wizard, then choose the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list.
350 Chapter 6
Note: You can plot no more than 2500 data points per curve, and you cannot plot more than four curves. 4. Click Finish to create the graph.
Figure 6-45 Intersections only appear for two our more curves, and a legend appears for each intersection.
1. Double-click the multiple area plot. The Graph Properties dialog box appears.
Figure 6-46 Using the Graph Properties Dialog Box to Identify Intersections
2. Click the Plots tab. 3. Select Area Fills from the Settings for list.
352 Chapter 6
4. Select Identify Intersections. 5. Click OK to close the dialog box and accept the changes.
1. Create an area plot. 2. Double-click the graph. The Graph Properties dialog box appears.
Figure 6-47 Using the Graph Properties Dialog Box to change the direction of the area fill
3. Click the Plots tab. 4. From the Settings for list, select Area Fills.
5. From the Direction drop-down list, select Up, Down, Left, or Right. 6. Click OK.
1. Double-click the area plot. The Graph Properties dialog box appears.
Figure 6-48 Using the Graph Properties Dialog Box to change the area fill color and pattern
2. Click the Plots tab. 3. From the Settings for list, select Area Fills.
354 Chapter 6
4. From the Color drop-down list, select (none) to create a transparent fill color, (Custom) to create a custom color, or an incremental color scheme to use a color array, or any one of many available colors. 5. From the Pattern drop-down list, select a pattern. 6. Click OK. For more information, see Changing Patterns and Fill Colors in Chapter 4.
Figure 6-49 An example of two plots, a bar chart and an area plot. In the area plot (in red), the area between the two curves is shaded.
You can shade the area between two curves by: Using the Object Properties dialog box to change the background color of the graph to match the lower shade. Using the Insert Graphic Cells dialog box to insert colors in to the worksheet, and then applying those to the plot.
To shade the area using the Object Properties dialog box:
1. Create an area plot that uses either X Many Y or XY Pairs data formats. Make sure, when in the Graph Wizard, that you first select to plot the column with the largest Y values for the upper curve. Then use the column with the smallest Y values for the lower curve.
356 Chapter 6
Figure 6-50 To shade between the curves, first create an area plot that uses the larger values for the upper curve.
2. Once you have created the graph, right-click. On the shortcut menu, click Object Properties. The Object Properties dialog box appears. 3. Click the Fill tab. 4. Under Fill Color, from the Color drop-down list, select a color that matches the color of the lower curve.
Figure 6-51 Shading the Area Using the Object Properties Dialog Box
5. Click Close. The graph appears with the area between the two curves shaded.
358 Chapter 6
Figure 6-52 The area between the two curves appears shaded, while the area under the lower curve matches the background..
For more flexibility you can define the area colors by inserting colors into a column in the worksheet and then use the front area color as the graph background color. 1. Create an area plot. 2. View the worksheet, and select a cell in the first row of an empty column. 3. From the menus select:
Insert Graphic Cells
4. Click the Colors tab. 5. Double-click to select two colors. In the first cell (row 1), select the color that you want the area to be and in the second cell (row 2), select the color you want the background to be.
Figure 6-53
6. Click Close to close the dialog box. 7. To assign the area plot colors to those in the worksheet, double-click the graph. The Graph Properties dialog box appears. 8. Click the Plots tab. 9. Select Area Fills. 10. Under Fill Color, scroll to the bottom of the Color drop-down list and select the column that contains the colors you selected in the Insert Graphic Cells dialog box.
360 Chapter 6
Figure 6-54 Select the column that contains the graphic cells that you inserted into the worksheet.
11. Click OK to apply the changes and close the dialog box. The graph now appears with the two shaded areas filled with the colors you inserted in worksheet; however, the background of the graph is still white.
Figure 6-55 Once youve selected the color for the lower curve, you still must match a color for the background.
12. Right-click the graph, and select Object Properties. The Object Properties dialog box appears.
362 Chapter 6
Figure 6-56 Selecting a color that matches the lower shaded area on the graph.
13. Click the Fills tab. 14. Under Fills color, from the Color drop-down list, select the color that matches the lower shaded area on the graph. 15. Click OK to apply the changes and close the dialog box. The graph appears with one shaded area between the two curves.
Figure 6-57 As in the example above, the graph appears with the background color matching the color of the lower shaded region.
Bubble Plots
Bubble plots are XY scatter plots that use symbols to represent not only XY locations, but also a third dimension represented by the size of the symbol. Use bubble plots to plot population density, epidemiological data, or other similar data sets where a third variable can be clearly illustrated by the size of the symbols.
364 Chapter 6
2. On the Standard tool bar, click the Graph Wizard button. The Graph Wizard appears. 3. From the Graph Types scroll-down list, select Bubble Plot, and click Next. 4. From the Data Format list, select the appropriate format, and click Next. 5. When you have selected all the columns to plot, including the Bubble Size column, click Finish.
Vector Plots
A vector plot is a line plot where the line represents the magnitude, and an arrow head represents the direction.
Figure 6-58
366 Chapter 6
3. From the Graph Types scroll-down list, select Vector Plot, and click Next.
Figure 6-60 Selecting a Vector Plot Data Format
The Create Graph - Data Format dialog box appears with two available formats: XYXY. Select XYXY when you want both Start and End points of the graph appear as XY coordinates. XYAM. Select XYAM if you want the end points to be represented in terms of angle and magnitude. If you select XYAM, when you click Next, the Create Graph - Angle Units panel appears.
Figure 6-61 Selecting Vector Plot Angle Units for the XYAM Data Format
Note: You may need to scale the magnitude data if you are using the XYAM format and your vector magnitudes are either much larger or smaller than the X,Y coordinates. The vectors should be large enough to be clearly visible, but also small enough to be spaced within the X,Y coordinates so that they do not overlap. If, for example, the magnitudes are too large, use a Quick Transform to divide the magnitude column by a constant that is roughly the ratio of the largest magnitude to the smallest X and Y coordinate increment. Then you can easily adjust this constant to obtain a pleasing graph. For more information, see Performing Quick Transforms in Chapter 13 4. Click Next. The Create Graph - Select Data dialog box appears.
368 Chapter 6
5. When you have selected all the columns to plot, click Finish.
1. Select the vector plot on the graph and then double-click. The Graph Properties dialog box appears.
Figure 6-63 The Graph Properties Dialog Box Displaying the Settings for Vector Plots
2. Under Settings for, click Vector. 3. Use the Arrowhead options to set the arrowheads angles in degrees and their lengths.
370 Chapter 6
1. Right-click the plot, and on the shortcut menu, click Add New Axis. The Graph Wizard appears.
Figure 6-64 Using the Graph Wizard - Add Axis Dialog Box to Select the Plot for the New Axis
2. Select to create either a new X axis or Y axis for the specified plot. 3. Click Next.
Figure 6-65 Selecting to Create a New Y Axis for the Selected Plot
4. Select which side of the graph to add the new axis. You can add the new axis to the left, right, top, or bottom of the graph. Selecting an Offset location moves the new axis slightly to the side, top, or bottom of the original axis.
Figure 6-66
5. Click Finish to add the new axis according to the specified settings. The New axis appears on the graph, and the plot re-scales to reflect the new axis
Figure 6-67 Example of a Second Y Axis Added to the Graph for a Line Plot
372 Chapter 6
1. Right-click the graph, and on the shortcut menu, click Add New Plot. The plot type does not matter, so long as it is a 2D Cartesian plot. 2. Pick any data format. 3. Pick empty columns when prompted to select the data to plot. 4. Create an axis for this "dummy" plot at the desired location 5. Select the new axis, then use manual scaling to set the appropriate range and check interval for the new axis. This scale is often a linear transformation of the opposite axis scale, for example, a Celsius scale to a Fahrenheit scale.
Figure 6-68 The second temperature axis for the single plot was created by first creating a dummy plot, creating a Y axis for the dummy plot, then manually scaling the axis range.
For more information, see Modifying Axes, Tick Marks, and Grids in Chapter 9.
Chapter
Create 3D (XYZ) plots from many worksheet columns or column triplets. XYZ plots must have at least one plot, but can display many more plots, each with a different type and style. Graphs can be rotated and shaded added to enhance the height and depth of mesh and bar charts. This chapter covers: An overview of 3D plots. For more information, see Working with 3D and Contour Graphs above. Creating 3D scatter plots and 3D bar charts. For more information, see Creating 3D Scatter Plots and 3D Bar Charts below. Creating trajectory plots. For more information, see Creating Trajectory Plots below. Creating waterfall plots. For more information, see Creating Waterfall Plots below. Creating mesh plots. For more information, see Creating Mesh Plots below. Changing graph perspective, rotation, and shading. For more information, see Changing Graph Perspective, Rotation, and Shading below. 3D axis placement. For more information, see 3D Graph Axis Placement. Creating contour plots. For more information, see Creating Contour Plots below. Modifying contour plots. For more information, see Modifying Contour Plots below.
373
374 Chapter 7
Mesh Plots
Mesh plots graph 3D data as a continuous surface with a mesh. Use the Graph Properties dialog box to modify mesh lines, color, transparency, and to enable the light source for shading. For more information, see Modifying Mesh Lines and Fill Color below.
Figure 7-2 Mesh Plot with No Fill Color and with a Gradient of Colors
3D Bar Charts
Create bar charts in 3D space using 3D data. Modify 3D bar charts by changing fill color and pattern. For more information, see Changing Patterns and Fill Colors in Chapter 4.You can also adjust bar width and spacing. For more information, see Changing Bar and Box Widths and Spacing in Chapter 4.
376 Chapter 7
Waterfall Plots
Waterfall plots graph 3D data as stacked line plots along the Y axis. Use the Graph Properties dialog box to modify plot lines, color, and transparency.
Figure 7-4 Waterfall Plots
3. From the Data Format list, specify how your data is formatted. The data formats available depend on the graph type you are making. 4. Click Next.
378 Chapter 7
Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the dialog box. 5. Click Finish. Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph. For more information, see Modifying Graphs in Chapter 4.
3. Click Finish. Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph.
380 Chapter 7
3. From the Data Format list, choose the appropriate data format. 4. Click Next.
Figure 7-10 Graph Wizard Select Columns Panel
Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the Selected Columns list. 5. Click Finish. Use the Graph Properties dialog box to modify plot lines, color, and transparency. For more information, see Modifying Mesh Lines and Fill Color below.
382 Chapter 7
3. From the Data Format list, choose the appropriate data format, and click Next.
Figure 7-12 Using the Create Graph Dialog Box to Pick Columns to Plot
Since you already selected columns prior to opening the Graph Wizard, your choices automatically appear in the dialog box. 4. Click Finish. Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot, or to add another plot to your graph. For more information, see Modifying Mesh Lines and Fill Color below.
Figure 7-13 Graph Properties Dialog Box Plots Tab Mesh Settings
2. Click the Plots tab. 3. From the Settings For list, select Mesh. 4. To change the color of the mesh, under Fill Colors, from the Color drop-down list, select a color. Select (none) to create a transparent mesh, select (Custom) to create a custom color, and select one of the color schemes or color columns to increment the mesh from bottom to top using a color array. For more information, see Using Custom Symbol, Fill, Line, and Color Increments in Chapter 4. 5. To make your mesh translucent, under Fill Colors, select Transparent. Objects behind it will be visible. Use this option to more clearly show the intersections between two or more 3D meshes. Tip: Set your display to High Color (16 bit) or True Color (24 bit) for this feature to work properly. Check your systems color capabilities under the Windows Display Properties Settings.
384 Chapter 7
6. If you are using a color scheme, under Fill Colors, from the Transition drop-down list, specify how the colors flow across the grid. Select Discrete to use an increment with a clear shift between colors, or select Gradient to use an increment with a gradual shift between colors. Note: The Transition drop-down list is available only when using a fill color scheme. 7. To change mesh lines, from the Settings For list, select Lines. Use the Color dropdown list to change line color. Selecting (none) creates transparent mesh lines, and selecting (Custom) enables you to use or create a custom color. For more information, see Using Custom Colors in Chapter 5. 8. To change line thickness, move the Thickness slider, or type a new value in the Thickness box. 9. Click OK.
2. Click the Graph tab. 3. From the Settings for list, select Rotation. This tab displays a Preview that shows how the current settings affect the selected graph. 4. To rotate the graph, move the Horizontal View and Vertical View sliders, or type horizontal or vertical values into the boxes.
386 Chapter 7
Note: Horizontal and vertical values are in degrees. Rotate the graph horizontally from 0 to 360, or vertically from 90 to +90. The recommended Horizontal View is 205, and the Vertical View is 25. The three solid red axes displayed in the Preview box of the 3D View tab are the origin axes for the rotation, and are used as reference when determining the angles of rotation. The rotation is displayed in the axes degrees from 0. The origin used to determine the degree from the horizontal or vertical is the intersection of the three axes. When both rotation angles are set to 0, the origin as you see the graph, is the left bottom rear corner. Note: The origin axes are not related to the axes marked with ticks and tick labels, but act as the zero point for tick labels and data.
Figure 7-15 A 3D graph with a horizontal rotation of 0, a vertical rotation of 0, and a perspective of 20.
5. To change the perspective of the graph, move the Perspective slider, or type a new value into Perspective box.
Figure 7-16 A 3D graph with a horizontal rotation of 0, a vertical rotation of 45, and a perspective of 20.
Figure 7-17 A 3D graph with a horizontal rotation of 45, a vertical rotation of 45, and a perspective of 20
Note: The Perspective value is based on the depth of the graph. A perspective of 0% means that the graph has no depth; 100% means that the graph has maximum depth. The recommended perspective is 20%.
388 Chapter 7
6. To enable the light source and create shading on your graph, select Enable Light Source. If the check box is cleared, the light source is not applied to the graph. Note: Set your display to High Color (16 bit) or True Color (24 bit) for this feature to work properly. You may check your systems color capabilities under the Windows Display Properties Settings.
Figure 7-21 The graph on the right has the light source option selected.
3D line and scatter plots are not affected by the light source option. 7. To return to the 3D View settings you had before applying any changes, click Revert to original settings. 8. Click OK.
390 Chapter 7
For more information, see Changing Graph Perspective, Rotation, and Shading above.
2. Click the Graph tab. 3. From the Settings for list, select Frame Lines.
4. From the Frame Lines drop-down list, select either: Relative to Viewer: If the frame is oriented from your perspective, one set of lines is composed of the three cube edges closest to you, and the other lines are the remaining sides of the cube. The position of these lines is independent of the graphs rotation. This is the default position. Relative to Graph Origin: If the frame is drawn according to the origin, one set of the lines is drawn over the origin axes, and the other lines draw the remainder of the cube. The position of these lines is dependent on the graphs rotation
Figure 7-23 These graphs use the Viewer as the point of reference. The graph on the left draws only the front lines, and the right graph draws only the back lines.
Figure 7-24 These graphs use the Origin as the point of reference. The graph on the left draws only the origin lines, and the right graph draws only the non-origin lines
392 Chapter 7
5. Hide frame lines, or add frame lines to your graph by selecting or clearing the appropriate Show check box. Selected frame lines are drawn. A graph cannot display frame lines for both the Relative To Viewer and Relative To Graph Origin perspectives. If Relative To Graph Origin is selected from the Frame Lines drop-down list, the Show check boxes for Relative To Viewer are cleared automatically, and vice versa. 6. To change the frame line type, under Front lines, from the Line Type drop-down list, select a line type. 7. To change a frame line color, under Front Lines, from the Color drop-down list, select a frame line color. 8. Select (None) from either list to create transparent frame lines. Choose (Custom) from the Color drop-down list to use or create a custom color. For more information, see Using Custom Colors in Chapter 5. 9. To the modify frame line thickness, move the Thickness slider, or type a new thickness value into the thickness field. 10. Click OK.
3. From the Data Format list, select the appropriate data format, and click Next. The Graph Wizard prompts you to specify which worksheet columns correspond to the data for your plot. Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in selected column list.
Figure 7-25 Graph Wizard Select Columns Panel
Selected Columns list.
Note: If you made a mistake picking data, click the wrong entry in the Selected Columns list, then select the correct column from the worksheet. You can also clear a column assignment by double-clicking it in the Selected Columns list. 4. Click Finish.
394 Chapter 7
Figure 7-26 Graph Properties Dialog Box Plots Tab Contours Settings
2. Click the Plots tab. 3. From the Settings For list, select Contours. 4. To modify contour lines, from the Contours drop-down list, select Major or Minor. The Line Styles reflect the contour you select in the Contour drop-down list. Select Major to change the Line Styles for major contours. Select Minor to change the Line Styles for minor contours. 5. To specify the line type of major and minor contour lines, from the Type drop-down list, select a line type. Select one of the incrementing schemes to increment contour line types, or select (none) to create transparent lines. 6. To select the color of the contour lines, from the Line Style Color drop-down list, select a color. You can choose from several predefined color schemes, or select (none) to create transparent lines. Select (Custom) option to create a custom color. 7. To set the thickness of the contour lines, move the Thickness slider, or type a new value in the Thickness box.
396 Chapter 7
8. Click OK.
2. Click the Plots tab. 3. From the Settings for list, select Contours. 4. From the Contours drop-down list, select Major. 5. From the Color drop-down list, select from several predefined color schemes. 6. From the Fill Start drop-down list, set the direction of the contour fills. The default direction is bottom. That is, the fill starts from the lowest z value.
You can also create filled contour plots automatically when you first create the graph. You can either select the Filled Contour Plot style from the 3D Graph toolbar, or choose Filled Contours from the Graph Wizard.
2. Click the Plots tab. 3. From the Settings for list, select Contours. 4. From the Contours drop-down list, select Minor.
398 Chapter 7
5. Under Fills, from the Color drop-down list, select (none). 6. Click OK.
2. Click the Plots tab. 3. From the Settings for list, select Scale.
4. From the Scale Type list, select Linear or Log (Common) scale. The linear scale uses a standard base 10 numeric scale, and the log scale uses a base 10 logarithmic scale. 5. To manually set the Z axis range, in the Start and End boxes, enter beginning and ending range values. 6. To automatically set the Z axis range, from the Calculation drop-down lists, select Data Range. SigmaPlot automatically determines the vertical range based on the Z data plotted. 7. To add padding to both ends of the axis, select Pad 5%. 8. To extend the range to the nearest major tick mark, select Nearest Tick. 9. Click OK.
400 Chapter 7
2. Click the Plots tab. 3. From the Settings for list, select Scale. 4. From the Apply to drop-down list, select the Major or Minor lines to modify. 5. Under Line intervals, from the Lines drop-down list, select one of the following intervals: Automatic: SigmaPlot automatically determines the interval at which contour lines are drawn. Manually: Manually set the number of contour lines are drawn. Enter the z interval in the Every field, and the value at which the first interval is drawn in the From field. Columns: Select the column used to determine major contour line z values.
Note: When major contour lines are plotted from a column, no minor lines are drawn. 6. Click OK.
1. Double-click the contour plot. The Graph Properties dialog box appears.
Figure 7-31 Graph Properties Dialog Box Plots Tab Labels Settings
2. Click the Plots tab. 3. From the Settings for list, select Labels.
402 Chapter 7
4. To display or hide contour labels, under Contour Labels, select or clear Major Contour Labels and Minor Contour Labels. 5. Selected options display labels, and cleared options hide labels. 6. To align contour labels parallel to the contour line, under Label Appearance, select Align With Contour Line. 7. Clear the option to align the contour labels parallel to the X axis. 8. To control how many labels appear for the contour lines, move the Label Frequency slider. 9. Move the slider towards Fewer to reduce the number of contour labels, or move the slider towards More to increase the number of contour labels. 10. To add to the contour labels, under Add to Major Labels and Add to Minor Labels, in the Prefix and Suffix boxes, type the prefix or suffix. 11. To separate a suffix or prefix from the tick label, type a space before a suffix or after a prefix. 12. Click OK.
2. Click the Plots tab. 3. From the Settings for list, select Details. 4. From the Type drop-down list, select Numeric, then use the Contour label appearance options. 5. From the use drop-down list, specify which type of numeric display to use. The Scientific Notation and Engineering Notation options always use scientific notation or engineering notation to display numbers. For large numbers options, use scientific or engineering notation only when numbers exceed a specified range. Use the Above and Below lists to specify the range beyond which scientific notation or engineering notation is used.
404 Chapter 7
For linear scale, you can always use scientific notation, or only when needed. If you use scientific notation only when needed, set the range to by typing values in the Lower and Upper ranges in the edit boxes. These values are expressed in log units. 6. Use the Precision options to specify the number of places used to display numeric tick labels. Select Automatic to let SigmaPlot automatically determine precision, or select Manual, then select the number of decimal places to use from the Places drop-down list.
2. Click the Plots tab. 3. From the Settings for list, select Details. 4. From the Type drop-down list, select Series.
5. From the Series list, select one of the following series: Months Days of the week Years Alpha Series Numeric 6. From the Length drop-down list, select the number of characters to use for the labels. 7. From the Start At drop-down list, select the series item to begin labeling tick marks with, then from the Step By drop-down list, select the frequency or increment for the series. 8. To restart tick labeling from a specified point, use the After and Repeat From dropdown lists. 9. Click OK to close the dialog box and make the changes.
406 Chapter 7
3. Click the Plots tab. 4. From the Settings for list, select Details. 5. From the Type drop-down list, select the column containing tick labels. 6. Click OK to close the dialog box and make the changes.
4. Click Contour label font to open the Text Properties dialog box. For more information, see Formatting Text in Chapter 5.
408 Chapter 7
Chapter
This chapter describes basic procedures specific to pie charts, polar plots, and ternary plots, including: Creating pie charts. For more information, see Pie Charts below. Changing pie chart slice settings. For more information, see Modifying Pie Charts below. Creating polar plots. For more information, see Polar Plots below. Modifying polar plots. For more information, see Modifying a Polar Plot below. Creating ternary graphs. For more information, see Ternary Graphs. Changing basic ternary graphs attributes. For more information, see Modifying Ternary Graphs.
Pie Charts
Pie charts plot a single worksheet column by representing each data point in the column as a pie slice. Each data point in the column is graphed as a slice size equivalent to the data points percent of the sum of all the data. The first pie slice starts at 0 degrees (3 oclock) by default. Additional slices are added counterclockwise, in the order the data points occur in the column.
409
410 Chapter 8
3. Specify which worksheet column corresponds to data for your plot. Since you selected a column prior to opening the Graph Wizard, your choice automatically appears in the dialog box and you can click Finish to create the pie chart. 4. If you selected the incorrect columns to plot, select a column either by clicking the corresponding column directly in the worksheet, or selecting the appropriate column from the Data for Pie list. Note: If you make a mistake while picking data, click the wrong entry in the Graph Wizard, then select the correct column from the worksheet. 5. Click OK. Use the Graph Properties dialog box to modify the pie chart, or reopen the Graph Wizard to pick a different data column for your plot. Note: You cannot add plots or axes to pie charts.
1. Double-click the pie chart. The Graph Properties dialog box appears.
412 Chapter 8
Figure 8-2 Graph Properties Dialog Box Plots Tab Pie Slices Settings
2. Click the Plots tab. 3. Select Pie Slices from the Settings For list. 4. Move the Counterclockwise from slider to change the starting angle. Increasing the starting angle for the first slice moves the starting slice counterclockwise. The default is 0 (3 oclock). 5. Click OK.
1. Double-click the pie chart. The Graph Properties dialog box appears.
Figure 8-3 Graph Properties Dialog Box Plots Tab Pie Slices Settings
2. Click the Plots tab. 3. Select Pie Slices from the Settings For list. 4. Select Single Slice from the Exploded Slices drop-down list. 5. Select the number of the slice to explode from the Slice drop-down list. 6. By default, the first slice begins at 0 and proceeds counterclockwise. If you have not rotated the pie chart, the slice number corresponds to the worksheet row number. 7. Click OK. Note: Choosing No Exploded Slices from the Exploded Slices drop-down lists replaces any exploded pie slices.
414 Chapter 8
1. Select an empty column. 2. Type a 1 in the same row as the data point for each row you want to emphasize with an exploding slice. 3. Double-click the pie chart. The Graph Properties dialog box appears.
Figure 8-4 Graph Properties Dialog Box Plots Tab Pie Slices Settings
4. Click the Plots tab. 5. Select Pie Slices from the Settings For list. 6. Select the column containing exploding slice data from the Exploded Slices drop-down list. 7. Click OK.
Polar Plots
Polar plots display data in the coordinate system format where r is the distance from the origin of the graph, and theta () is the angle between the positive horizontal axis and the radius vector extending from the origin to the plotted data point. Use polar plots to show data where one value (q) is periodic in nature, like a clock. An example of this is a graph that shows average temperatures of differing geographical regions during the days of a month, or months of a year.
Figure 8-5 Polar Plots
416 Chapter 8
3. If you want to create a polar plot with its angles increased in a clockwise direction, select Clockwise. This creates a plot where the Range Lower Bound starts at the top of the graph. On polar plot set to use a counter clockwise direction, Range Lower Bound starts at the 3 oclock position. 4. Click Next. 5. Select the appropriate data format from the Data Format list. 6. Click Next. 7. Click Finish. Use the Graph Properties dialog box to modify the plot, or reopen the Graph Wizard to pick different data columns for your plot. For more information, see Modifying Graphs in Chapter 4.
Ternary Graphs
Ternary graphs plot data on an XYZ coordinate system in the form of three variables that add up to 100% or 1. These variables are typically the normalized proportions of three substances and are plotted on three axes generally arranged as an equilateral triangle. These graphs are also commonly referred to as triangle plots.
418 Chapter 8
Figure 8-6 Selecting a Ternary Graph Data Format from the Graph Wizard
Since you selected columns prior to opening the Graph Wizard, your choices automatically appear in the dialog box. Tip: If you made a mistake picking columns, highlight the wrong entry in the Graph Wizard, then choose the correct column either in the worksheet or from the column list. 4. Click Finish.
Use the Graph Properties dialog box to modify the plot or to open the Graph Wizard to pick different data columns to plot or to add another plot to your graph. For more information, see Modifying Ternary Axes in Chapter 9.
420 Chapter 8
Chapter
Axes are the scales or rulers along which data is plotted. 2D Cartesian graphs have an X (horizontal) axis, and a Y (vertical) axis. For 3D graphs, the X and Y axes form the base of the graph, and the Z axis is the vertical axis. Polar plots use an angular axis to draw the circumference of the plot and the radial axes to draw the radius of the plot. An axis is always associated with at least one plot on a graph, and determines the scaling of the plot. Each axis consists of pairs of lines that you can move and modify independently. Tick marks are short lines along the axis that represent scale intervals. You can display and modify tick marks for each axis. Grid lines are attached to the graph planes, and can be drawn at tick intervals for all axes. Make most axis modifications using the Axes tab of the Graph Properties dialog box. This chapter covers: Changing axis scales and ranges. For more information, see Changing Axis Scales and Rangebelow. Changing scale type. For more information, see Changing Scale Type below. Hiding, displaying, and deleting axes. For more information, see Hiding, Displaying, and Deleting Axes below. Setting axis breaks. For more information, see Setting Axis Breaks below. Working with axis titles and tick labels. For more information, see Working with Axis Titles and Tick Labels below. Changing tick mark intervals. For more information, see Changing Tick Mark Intervals below. Changing tick mark appearance. For more information, see Changing Tick Mark Appearance below.
421
422 Chapter 9
Changing tick labels. For more information, see Changing Tick Labels below. Displaying grid lines and backplanes. For more information, see Displaying Grid Lines and Backplanes below. Modifying polar axes. For more information, see Modifying Polar Axes below. Modifying ternary axes. For more information, see Modifying Ternary Axes below.
Tip: You can also double-click the axis on the graph. The Graph Properties dialog box appears. 3. If necessary, click the Axis tab, and then select Scaling from the Settings for list. The available scale types appear under the Scale type list. They include: Linear. For more information, see Linear below. Common Log. For more information, see Common Log below. Natural Log. For more information, see Natural Log below. Probability. For more information, see Probability below. Probit. For more information, see Probit below. Logit. For more information, see Logit below. Category. For more information, see Category below. Date/time. For more information, see Date/Time below.
Linear
A linear scale is a standard base 10 numeric scale. (This scale is recommended for a date axis when an exact representation of spacing depicted by dates is not required. Otherwise, use the date/time scale.)
Common Log
A common log scale is a base 10 logarithmic scale.
Natural Log
A base e logarithmic scale.
Figure 9-1 Graphs of the Same Data Using Linear, Common Log, and Natural Log Scales
Probability
A probability scale is the inverse of the Gaussian cumulative distribution function. The graph of the sigmoidally shaped Gaussian cumulative distribution function on a probability scale is a straight line. Probabilities are expressed as percentages with the minimum range value set at 0.001 and the maximum range value set at 99.999. The default range depends on the range of the actual data.
Probit
A probit scale is similar to the probability scale; the Gaussian cumulative distribution function plots as a straight line on a probit scale. The scale is linear, however, with major tick marks at each Normal Equivalent Deviation (N.E.D. = X - )/) plus 5.0.
424 Chapter 9
At the mean (X = ) the probit = 5.0; at the mean plus one standard deviation (X = + ) the probit = 6.0, etc. The default range is from 3 to 7. The range limit for a probit axis scale is 1 to 9.
Figure 9-2 Graphs of the Same Data Using Linear, Probability, and Probit Scales
Logit
A logit scale uses the transformation logit=ln(ya-y) where a = 100
Category
A category scale uses numerical values or text from a worksheet column used to generate a plot. Each distinct entry in the column is a separate category against which the corresponding data values are plotted. This scale is commonly used for bar charts or other plots to graph different categories of data.
Any plot generated by plotting a column containing any text versus a column containing data will use a category axis automatically. For more information, see Using a Category Scale below. You can select a category scale for numeric data, and each unique value will be treated as its own category. Note: If a column contains more than one instance of the same category, the category appears only once, and all corresponding data is plotted within that category.
Date/Time
Date and time formatted data are automatically plotted using a Date/Time axis scale. This scale is specifically designed to handle true calendar date and time units, labeling and spacing. You can: Plot date and time data.For more information, see Entering Dates and Times in Chapter 3. Change date and time labels.For more information, see Changing Date and Time Tick Labels below. Change data and time intervals.For more information, see Tick Intervals for Date/Time Axes below. Although you can plot numeric data as date and time, you should first view these numbers as date and time in the worksheet to make sure they are sensible values.
426 Chapter 9
If a worksheet cell is a label, it will not plot as a date and time value. In this case, you may want to reenter the label as a date and time value.
Reciprocal
A reciprocal axis scale uses the multiplicative inverse of a number, where x is the number, which when multiplied by x, yields 1. Its equation is 1/x or x-1
Figure 9-4 An Arrhenius Plot with a Reciprocal Axis Scale
Researchers often use a reciprocal scale when creating graphs for reliability studies. For example, there is a theory that aging accelerates as temperature rises. The plot in this case would use a reciprocal temperature (1/T (1/mK)) x-axis and a log (ln k) yaxis. The slope of the line fitted through the plot is the activation rate of the studied chemical reaction, that is, the reaction causing the failure.
Weibull
The Weibull axis displays the Cumulative Percent Failure (CPF) using the Weibull distribution formula: y = ln(ln(1/(1-CPF/100))) This scale is frequently used for life data analysis. The Failure Time, typically a log scale on the x-axis is plotted against the Cumulative Percent Failure, typically on the y-axis.
1. Double-click the axis to modify. The Graph Properties dialog box appears.
Figure 9-5 Graph Properties Dialog Box Axes Tab Scaling Settings
428 Chapter 9
3. From the Axis drop-down list, select the axis you wish to modify. 4. From the Settings for list, select Scaling. 5. To automatically set the axis range, select Data Range from the Calculation list. SigmaPlot automatically determines the axis range based on the data plotted. For log axes, or axes that forbid zero or negative numbers, the minimum is set to the nearest major tick mark beyond the smallest value. 6. To manually set the axis range, select Constant then type beginning and ending axis range values in the Start and End edit boxes. Note: Date/Time axes display the ranges in date and time units. 7. Select Pad 5% to add padding to both ends of the axis. 8. Select Nearest Tick to extend the range to the nearest major tick mark. 9. Click OK.
1. Select the axis to modify. 2. On the Properties toolbar, click the Axis Scale button. The Graph Properties dialog box appears.
Figure 9-6 Using the Scale Type list from the Axes Tab of the Graph Properties Dialog Box
3. From the Settings for list, select Scaling. 4. From the Scale Type list, select the desired axis scale type. The default axis scale is Linear for all numeric data, Category for text data, and Date/Time for date and time data. 5. Click OK.
430 Chapter 9
1. Enter your category data (text) into a worksheet column, and corresponding data into adjacent worksheet columns. 2. On the Graph toolbar, click the graph type and style you want to create. The Graph Wizard appears. 3. Select the data format, and click Next. 4. Since you have not already selected your data from the worksheet, click the worksheet columns to assign them under Selected Columns. Plot your category column as the category axis data type, and pick your other column(s) as the corresponding data type. 5. Click Finish to create the graph.
To modify a plot to use a category scale:
1. Double-click the axis you want to modify. The Graph Properties dialog box - Axes tab appears.
2. Select Scaling from the Settings list. 3. Select Category from the Scale Type drop-down list. 4. Click Apply to change the scale type without closing the dialog box. 5. Click the Plots tab, and then click Graph Wizard. The Graph Wizard - Modify Plot dialog box appears. 6. Under Data Format, select the data format you want to use, and click Next. 7. Click the columns in the worksheet to choose the worksheet columns to assign to plotted data under Selected Columns. Plot your category column as the data type you want to use as the category axis, and pick you other columns(s) as the corresponding data type. 8. Click Finish to create the graph with the newly assigned worksheet data and modified axis.
1. Enter dates or times into a column of a worksheet. For example, enter 1/1, 2/1, 3/1, etc., indicating months and days. 2. Enter corresponding data into a separate worksheet column or columns. 3. Drag the pointer over both the date and data columns. 4. Create the graph using the Graph toolbar or the Graph Wizard. 5. Plot your date and time column as the date/time axis.
432 Chapter 9
6. Pick your other column(s) as the corresponding axis. 7. Click Finish to create the graph.
1. Double-click the axis (you can double-click hidden axes as well). The Graph Properties dialog box appears.
Figure 9-9 Graph Properties Dialog Box Axes Tab Line Settings
2. Click the Axes tab. 3. From the Settings for list, select Lines. 4. Under Show/Place Axes, select an axis to display that axis, or clear an axis to hide it. Hidden axes hide all ticks and labels associated with that axis.
434 Chapter 9
Note: You can hide 3D axes, but if frame lines are active, a line will remain present. For more information, see Drawing, Modifying, and Hiding Frame Lines in Chapter 7. Also, if the graph has grid lines, a line will remain present. For more information, see Displaying Grid Lines and Backplanes below. 5. Click OK.
1. Double-click the axis. The Graph Properties dialog box appears. 2. Click the Axes tab. 3. Select Lines from the Settings for list. 4. Select the axis you want to modify from the Axis drop-down list. 5. To change the color of the axis, under Line Properties, select a color from the Color drop-down list. Choose (None) to make the axis transparent, and choose (Custom) to open the Custom Color dialog box. 6. To change the thickness of the axis, under Line Properties, move the Thickness slider or type a thickness in the Thickness box. 7. Click OK to apply the changes and close the dialog box. Note: 3D graph frame lines are drawn over axes lines and may obscure 3D axes modifications.
Figure 9-10 The Line Tab of the Object Properties Dialog Box Note that the Type option is unavailable for axes.
Moving Axes
You can move 2D axes with your mouse, or to a precise location with the Graph properties dialog box. You cannot move 3D axes, but you can hide them from view. For more information, see Hiding, Displaying, and Deleting Axes above.
436 Chapter 9
1. Double-click the axis you want to move. The Graph Properties dialog box appears.
2. Click the Axes tab. 3. From the Settings for list, select Lines. 4. Under Show/Place Axes, move the sliders to change the percentage in the Top and Bottom boxes for X axes or Y axes, or type the value in the fields. 5. Locations are described as the percentage of the graph dimension the axes lie from the original position. To move an axis up or right, enter a percent greater than 0% (positive). To move an axis down or left, enter a percent less than 0% (negative). The defaults are 0.0%, and Normal. 6. Click OK.
438 Chapter 9
1. Double-click the axis where you want to add the break. The Graph Properties dialog box appears.
Figure 9-14 Graph Properties Dialog Box Axes Tab Breaks Settings
2. Click the Axes tab. 3. From the Settings for list, select Breaks. 4. From the Break Range group box, select Show Break. 5. In the Omit boxes, enter the Break to omit. 6. To specify the break position, move the Position slider. The location of the break is determined as a percent of the total axis length, from the origin. 7. To set the width of the space between break lines, move the Gap Width slider. 8. To set a post break interval, type a value in the Post Break Interval box. The default value is the interval specified for the axis range. Note: Tick values from a column are not applied to the post break interval.
440 Chapter 9
9. To set axis break properties, under Break Properties, from the Symbol drop-down list, select a break symbol type then use the Length, Color, and Thickness options. 10. Click OK.
1. On the graph page, double-click the axis title. The title appears highlighted. 2. Make your changes to the title. Note: You can also rename an axis title from within the Axis tab of the Graph Properties dialog box. Double-click the axis, then click Rename. The Edit Text dialog box opens.
1. Right-click the axis title. On the shortcut menu, click Edit. The Edit Text dialog box appears. 2. Select a degree of rotation for the selected label from the Rotation drop-down list.
1. Double-click the desired axis. The Graph Properties dialog box appears.
Figure 9-15 Graph Properties Dialog Box Axes Tab Labels Settings
2. Select Labels from the Settings for list. 3. To view or hide the axis title, under Show Axis Title select or clear Bottom or Top to specify the position of the axis label. 4. To view or hide Major Tick labels, from the Apply to list, select Major Ticks, then under Major Tick Labels, then select or clear Bottom or Top to specify the position of the tick label.
442 Chapter 9
5. To view or hide Minor Tick labels, from the Apply to list, select Minor Ticks, then under Minor Tick Labels select or clear Bottom or Top to specify the position of the tick label. 6. Click OK.
Figure 9-16 The Axes Major Tick Intervals Options for a Linear Axis
2. Click the Axes tab. 3. From the Settings for list, select Ticks. 4. To change tick intervals, select from the Ticks and Every drop-down lists in the Tick Intervals group box. 5. If you select Manual, enter interval values by typing into the Every and From fields. The value in the Every field specifies how often major tick marks appear, and the From value specifies the origin point on the axes from which major tick marks start appearing. For example, if you type 0 into the From field and the 2 into the Every field, the major tick marks appear at even numbers on the axis. If you type 1 into the From field , the major tick marks appear at odd numbers on the axis. Custom Tick Intervals: You can also choose major tick interval values from the worksheet from the Major Tick Intervals list. Custom tick intervals are not available for minor ticks. For more information, see Customizing Tick Intervals below.
444 Chapter 9
6. Click OK.
1. Change the axis scale to a log axis. For more information, see Changing Axis Scales and Range above. 2. Double-click the tick marks. The Graph Properties dialog box appears.
Figure 9-18 The Axes Minor Tick Intervals Options for a Log Axis
3. Click the Axes tab. 4. From the Settings for list, select Ticks. 5. From the Apply to drop-down list, select Minor Ticks. 6. Select all minor tick intervals you want to appear, and clear those you want hidden. 7. Click OK.
446 Chapter 9
Figure 9-19 Tick Intervals for Natural Log and Logit Scales
1. Double-click the tick marks. The Graph Properties dialog box appears.
2. Click the Axes tab. 3. From the Settings for list, select Ticks. 4. Under Tick Intervals, from the Density drop-down list, select a tick mark density. 5. Click OK.
1. Double-click the tick marks. The Graph Properties dialog box appears.
448 Chapter 9
2. Click the Axes tab. 3. From the Settings for drop-down list, select Ticks. 4. Under Tick Intervals, from the Type drop-down list, select a tick interval type. Tick intervals are defined by the unit type used and the selected count. Dates fall at 12:00 AM of the first day for that period. The major tick interval options available are limited by the data range. You are prevented from selecting time units that would create too many tick marks. To increase the period between tick occurrences, change the Count. For example, set ticks to occur every other Type date by changing the Count to 2, or every fifth by changing the count to 5. Counts must be positive integers. 5. To set minor tick intervals, from the Apply to drop-down list, select Minor Ticks. 6. Select the minor tick Type and Count. Any time unit equal to or smaller than the Major interval type can be used as the Minor interval type. Note: Do not select a minor interval that creates hundreds or even many thousands of minor tick marks.
1. Enter the desired tick marks into an empty worksheet column. 2. Double-click the tick marks. The Graph Properties dialog box appears.
3. Click the Axes tab. 4. From the Settings for drop-down list, select Ticks. 5. From the Apply To drop-down list, select Major Tick Intervals. 6. Under Major Tick Intervals, from the Ticks drop-down list, select the column number or title of the column you want to use for major tick marks. 7. Click OK. The numeric values of the intervals are automatically used for tick labels, that you can modify them like any other tick labels.
450 Chapter 9
Tick marks for polar plots. For more information, see Displaying and Changing Radial Axis Ticks and Labels below. Tick marks for ternary graphs. For more information, see Changing Ternary Axis Tick Marks and Tick Labels below.
1. Double-click the tick marks. The Graph Properties dialog box appears.
Figure 9-23 The Axes Tick Direction Options
2. Click the Axes tab. 3. From the Settings for list, select Ticks. 4. From the Direction list for either axis: Select Outward, to point tick marks away from the graph. Select Inward to point tick marks toward the inside of the graph.
Select Both to point tick marks in both directions. Select (none) to hide tick marks. Note: The options that appear under Direction are dependent upon what type of graph you are modifying. If you are modifying a 3D plot, then the options are Front and Rear. If you are modifying a polar plot with a radial axis, then the options are Outer and Inner. In a polar plot with a radial axis, you change the direction of the spokes. Directions for tick marks are independent for either side of the axis.
1. Click the tick marks on the page. 2. Press Delete, or right-click and from the shortcut menu, click Hide.
1. Double-click the tick mark. The Graph Properties dialog box appears.
452 Chapter 9
Figure 9-24 Graph Properties Dialog Box Axes Tab Ticks Settings
2. From the Settings for list, select Ticks. 3. From the Apply to drop-down list, select Major Ticks or Minor Ticks. 4. To change tick length an thickness, under Tick Line, move the Length and Thickness sliders. 5. Select a color from the Color drop-down list. Choose from any of the listed colors, or select (Custom) to use a pre-defined custom color or create your own color. For more information, see Using Custom Colors in Chapter 5. Select (none) to create transparent tick marks. 6. Click OK.
Both major and minor tick marks. Standard numeric labels. Time and series labels. You can also add a suffix or prefix to all major or minor tick labels on a selected axis, and modify the calculation and precision of numeric labels, view different dates and times, select among many different series labels, and change the font and other text attributes.
1. Right-click the tick labels, and from the shortcut menu, click Text Properties.
Figure 9-25 Selecting a Column for Tick Label Intervals
The Text Properties dialog box appears. 2. Click the Font tab. 3. Change text attributes for tick labels the same way you would for any text label. You can also use the Rotation drop-down list on the Paragraph tab to rotate tick labels, but no other Paragraph settings are applied to tick labels.
454 Chapter 9
1. Double-click the tick labels you want to change. The Graph Properties dialog box appears.
Figure 9-26 Changing the Tick Label Type
2. Click the Axes tab. 3. From the Settings for list, select Tick Labels. 4. From the Apply to drop-down list, select either Major Ticks or Minor Ticks.
5. To use a numeric type of tick label, from the Type list, select Numeric, then use the Label Notation options. 6. To use a series type of tick label, from the Type list, select Series. 7. Click OK. Note: If you want to plot data versus true calendar dates, you should have entered date and time data in the worksheet, and use a date/time axis scale.
1. Double-click the tick labels of the axis you want to change. The Graph Properties dialog box appears.
Figure 9-27 Selecting Numeric Tick Label Notation
456 Chapter 9
3. From the Settings for list, select Tick Labels. 4. From the Apply to drop-down list, select either Major Ticks or Minor Ticks. 5. Under Label appearance, from the use drop-down list, select the type of label notation to use. Scientific Notation or Engineering Notation for large numbers use scientific or engineering notation only when numbers exceed a specified range. Use the When below and or above drop-down lists to specify the range beyond which scientific notation or engineering notation is used. Once a label exceeds the range, then all the labels will use the specified notation. For log axes, you can select to display the number, only the Exponent, or both the Base and exponent. For linear axes, you can select Scientific notation or Engineering notation to use always, or you can select Scientific notation, for large numbers or Engineering notation, for large numbers to use only when needed for large numbers. To specify when scientific notation is needed, enter the lower and upper ranges in the When below and or above.
Figure 9-28 Log Scale Y Axes Using Numbers, Exponent Only, and Base and Exponent
6. To divide numeric tick label values by a specific number, enter a divisor in the Factor Out drop-down list. A value of 2 divides label values in half, a factor of 0.5 doubles the tick label values, etc. 7. To specify the number of places used to display numeric tick labels, under Precision, select Automatic to let SigmaPlot automatically determine precision, or select Manual, then select the number of decimal places to use from the Places drop-down list.
8. Click OK.
1. Double-click the tick labels of the axis you want to change. The Graph Properties dialog box appears.
Figure 9-29 Selecting Series Tick Label Format
2. From the Settings for list, select Tick Label. 3. From the Apply to drop-down list, select either Major Ticks or Minor Ticks. 4. From the Type drop-down list, select Series. 5. From the Series drop-down list, select a series. 6. From the Length drop-down list, set the number of characters to use for the tick label.
458 Chapter 9
7. From the Start At drop-down list, specify which series item to begin labeling tick marks with. 8. From the Step By list, set the frequency, or increment, of series items to use. For example, if you are using a Days of the Week series, you might choose to start with Monday, and to step labeling by 2 days at a time. Tick labels appear as Monday, Wednesday, Friday, Sunday, Tuesday, etc. 9. To re-start tick labeling from a specified point, use the After and Repeat From dropdown lists. For example, if you were using a Days of the Week series, and were stepping by 2 days at a time, you might use the After and Repeat From lists to specify that after Friday, repeat the series from Monday. Tick labels appear as Monday, Wednesday, Friday, Monday, Wednesday, Friday, etc. 10. Click OK.
1. Double-click the axis you want to change. The Graph Properties dialog box appears. 2. Click the Axes tab. 3. From the Settings for list, select Labels. 4. From the Apply to drop-down list, select either Major Ticks or Minor Ticks. 5. To add a prefix or suffix to the major or minor tick labels, type the prefix or suffix into the appropriate Add to Tick Labels Prefix or Suffix boxes. All labels on the selected axis appear with the specified suffix or prefix.
You can use any keyboard or extended characters. Use the Windows Character map accessory program, or Alt+Numeric keypad combinations to enter extended characters like degrees symbols (Alt+0176). 6. Click OK.
1. Double-click the tick labels of the axis you want to change. The Graph Properties dialog box appears.
Figure 9-30 Changing Date/Time Tick Labels
2. From the Settings For list, select Tick Label. 3. From the Apply to drop-down list, select either Major Ticks or Minor Ticks.
460 Chapter 9
4. To change the display Date format, select a format from the list, or use the following table to enter a new label, using any additional characters as delimiters (i.e., slashes, commas, spaces, etc.). As you enter a different format, the Sample window shows an example of the label. Typing: Displays: M/d/yy No leading 0 for single digit month, day or year MM/dd/yy Leading 0 for single digit month, day or year MMMM Complete month dddd Complete day of week yyy or yyyy Complete year MMM Three-letter month ddd Three-letter day of week gg Era (AD or BC)
5. To change the display Time format, select a format from the list, or use the following table to enter a new label, using any additional characters as delimiters (i.e., colons, spaces, etc.). As you enter a different format, the Sample window shows an example of the label. Typing: Displays: hh or h 12 hour clock HH or H 24 hour clock mm or m Minutes ss or s Seconds uu or u Milliseconds H:h:m:s:or u No leading zeroes for single digits HH:hh:mm:ss:uu Leading zero for single digits tt Double letter AM or PM
462 Chapter 9
1. Enter the labels you want to use in a worksheet column in the order you want them to appear. Enter minor labels in the right adjacent column.
Figure 9-31 Tick Labels from a Column using the Symbol Font
Note: To skip specific labels, leave an empty cell for that tick mark when entering the labels into the worksheet column.
2. Double-click the axis tick labels you want to modify. The Graph Properties dialog box appears.
Figure 9-32 Selecting a Columns as the Source for the Tick Labels
3. Click the Axes tab. 4. From the Settings for list, select Tick Labels. 5. From the Type drop-down list, select the column to use for major labels. Labels for minor ticks are automatically taken from the column to the right of the major tick labels. 6. To change the font used for the tick labels, click Font. The Text Properties dialog box appears. You can use the Symbols font for Greek characters, and the Wingdings and other symbol fonts for iconic labels. 7. Click OK.
464 Chapter 9
Modifying Backplanes
To change back planes:
1. Double-click the graph to modify. The Graph Properties dialog box appears. 2. Click the Graph tab. 3. Under Settings for, select Backplanes. 4. If your graph is a 3D graph, from the Plane list, select the plane to modify. Note: When modifying a 2D graph, only one plane is available. 5. To select a background color for the selected plane, under Background, select a color from the Color drop-down list. 6. Select any of the listed colors, or select (Custom) to use or create a custom color. For more information, see Using Custom Colors in Chapter 5. 7. Select (none) to create a transparent plane. Transparent planes are especially useful when superimposing graphs over one another. The grid lines available for Cartesian plots are X, Y, and Z for 3D plots. The grid lines for polar plots are for the Angular and Radial axes. Ternary plots have X, Y and Z direction grid lines. 8. Click OK.
1. Double-click the graph to modify. The Graph Properties dialog box appears. 2. Click the Graph tab. 3. Under Settings for, select Grid Lines.
Figure 9-33 Selecting the Grid Lines
4. To change grid line thickness, under Gridlines, move the Thickness slider, type a thickness value in the Thickness box. 5. To change grid line style, under Lineproperties, from the Style drop-down list, select a style. 6. To change grid line color, under Line properties, from the Color drop-down list, select a color. Choose any of the listed colors, or choose (Custom) to use or create a custom
466 Chapter 9
color. Choose (none) to turn off grid lines. For more information, see Hiding and Viewing Grid Lines below. 7. To move the grid behind or in front of the plot, from the Layering drop-down list, select to move either the plot or grid to the front. This feature is especially useful for bar charts, and is not available for 3D plots. 8. Click OK.
Figure 9-34 A Bar Chart with a White Backplane and White Grid Lines Placed in Front of the Plot
1. Open the Graph Properties dialog box. 2. Click the Graph tab. 3. Select Grid Lines from the Settings for list. 4. To hide grid lines, under Style, select (none) from the Style drop-down list. 5. To display grid lines, change the style to a style other than (none).
6. Click OK.
Angular Axes
You can draw angular axes along the inner and outer circumferences of the graph. By default, the inner axis is not displayed. You can modify angular axes by: Changing axis titles. For more information, see Working with Axis Titles and Tick Labelsabove.
468 Chapter 9
Displaying or hiding either axis. For more information, see Hiding, Displaying, and Deleting Axes above. Changing axis lines. For more information, see Changing Axis Line, Color, and Thickness above. Changing axis scaling, range, and rotation. For more information, see Changing Angular Axis Scaling and Position below. Changing tick marks. For more information, see Changing Tick Mark Intervalsabove. Changing axis tick labels. For more information, see Changing Tick Labels above.
Figure 9-36 Graph Properties Dialog Box Axes Tab Scaling Settings
2. Click the Axes tab. 3. Select Scaling from the Settings for list. 4. To change the axis scale used, choose the desired axis scale type from the Scale Type list. For more information, see Axis Scale Types above. 5. To change the measurement units of the angular axis, select measurement units from the Angular Axis Units drop-down list. If you do not see the axis units you want to use for your polar plot listed in the list, select Other, then type new axis range values in the Range Lower Bound and Range Upper Bound fields. If using a predefined measurement unit, the Range Lower Bound and Range Upper Bound box values are entered automatically. For more information, see Using a Custom Axis Scale above. Note: The only effect of changing units is to change the pre-defined axis range. This range can be manually changed regardless of the current units. 6. To change the size of the displayed arc of the polar plot, move the Arc slider. A setting of 360 degrees displays the entire circle, 270 degrees displays three-quarters of the circle, and 90 degrees displays half of the circle.
470 Chapter 9
Note: If you change the arc of the angular axis, the axis range remains the same. The current axis range appears along the new distance of the arc. 7. To change the start angle for the displayed arc, move the Start Angle slider. The default is 0 degrees (3 oclock). Rotation is counterclockwise.
Figure 9-37 Polar plots with: Starting angle of 315 and arc of 270; start angle of 0 and arc of 180; and start angle of 135 and arc of 22.5.
8. Click OK.
Figure 9-38 Graph Properties Dialog Box Axes Tab Lines Settings
2. Click the Axes tab. 3. Select Lines from the Settings for list. 4. To change the percentage in the Outer and Inner axes, under Show/Place Axes, move the Outer and Inner slider controls. Locations are described as the percentage of the distance the axes lie from the center of the graph. To move an axis out, increase the percent. To move an axis in, decrease the percent. 5. Click OK.
Radial Axes
The radial axes are drawn along the radius of the graph, and by default are displayed as four axes extending from the center of the graph to the outer of edge the graph. Each of the radial axes is a representation of the same data, so the range and scale must be
472 Chapter 9
the same for each radial axis; however, you can modify the color, tick marks, labels, location, and display of each radial axis independently. Modify radial axes by: Displaying or hiding any axis. For more information, see Hiding, Displaying, and Deleting Axes above. Changing display of axis and tick label titles. For more information, see Displaying and Changing Radial Axis Ticks and Labels. Changing axis lines. For more information, see Modifying Radial Axes Lines and Position below. Changing axis scaling. For more information, see Changing Axis Scales and Range above. Changing tick marks. For more information, see Changing Tick Mark Intervals above. Changing axis tick label type. For more information, see Changing Tick Labels above.
1. Double-click the graph to open the Graph Properties dialog box. 2. Click the Axes tab. 3. Select Lines. Moving a Radial Axis.
1. Select the axis on the page. 2. Use the mouse to drag it to a new location. Radial axes rotate about the center of the graph like the spokes of a wheel. Setting Radial Axis Positions to Exact Degree Positions.
To set radial axis positions to exact degree positions
2. Click the Axes tab. 3. Select Lines from the Settings for list. 4. To move a radial axis, under Show/Place Axes, move the sliders to set a new location. The axis location is in degrees from 0 degrees (3 oclock). The defaults are 0 degrees, 90 degrees, 180 degrees, and 270 degrees.
474 Chapter 9
5. To offset an axis from the center of the graph, move the Axes Start slider to change the length of the radial axes. Setting the slider to 0% draws the axis from the center of the graph outward, 25% draws the axis beginning a quarter of the distance from the center, 50% draws it half the distance from the center, and so on.
Figure 9-40 Radial Axes in the Default Positions, and Offset by 45 degrees with an Axes Start of 30%.
Figure 9-41 Graph Properties Dialog Box Axes Tab Lines Settings
2. Click the Axes tab. 3. Select Lines from the Settings for list. 4. To view or hide a radial axis, select Spoke 1, 2, 3, or 4. 5. Select or clear the Show/Place Axes boxes to show or hide the axis. 6. To change line color and thickness, under Line Properties, select a color and thickness from the Color and Thickness drop-down lists. 7. Click OK.
476 Chapter 9
Other than display and position, polar plot tick marks and labels have the same options as Cartesian graph tick marks and labels.
Viewing, Hiding, or Moving Titles and Tick Labels on the Radial Axes
To view, hide, or move titles and tick labels on the radial axes:
2. Click the Axes tab. 3. Select Labels from the Settings for list. 4. Select either Minor Ticks or Major Ticks from the Apply to drop-down list. 5. To move or hide the major or minor tick labels on the radial axes, use the Major (or Minor) Tick Labels options.
6. Select (none) to hide the labels. 7. Select clockwise or counterclockwise to move the label from one side of the axis to the other. 8. Click OK.
Specifying the Direction for Radial Axis Tick Marks for Each Pair of Radial Axes
To specify the direction for radial axis tick marks for each pair of radial axes:
1. Double-click any radial axis tick mark. The Graph Properties dialog box appears.
Figure 9-43 Graph Properties Dialog Box Axes Tab Ticks Settings
478 Chapter 9
2. Click the Axes tab. 3. Select Ticks from the Settings for list. 4. Select either Minor Ticks or Major Ticks from the Apply to drop-down list. 5. Use Direction options to change the tick directions on the radial axes. You can only change the directions for Spokes 1 and 3 together, and for 2 and 4 together. Note: Selecting Inward orients the ticks clockwise, and Outward points the ticks counterclockwise.
Figure 9-44 Polar Plots with All Ticks Pointing Inward, Spokes 1, 3 Inward and 2,4 Outward, and All Ticks Pointing in Both Directions
6. Selecting Both directions draws ticks both clockwise and counterclockwise, and selecting (none) hides the tick marks. 7. Click OK.
1. Double-click the axis. The Graph Properties dialog box appears. 2. Click the Axes tab. 3. Select Labels from the Settings for list. Tip: To identify which axis is associated with and axis title, keep in mind that the title at the apex is always at the 100% point or maximum for that axis. 4. Under Show Axis Title, select the desired Axis title location from the At drop-down list. 5. Select Axis Title in the Rotate with Axis group box to rotate the axis title parallel to the axis. 6. Click Apply. 7. Continue to modify the titles of the other axes. Specify the axis title you want to change using the Axis list, then make the desired changes. 8. When you have finished, click OK.
480 Chapter 9
Figure 9-45 The titles along the axes are also rotated with the axes.
A selected apex displays a black, square selection handle and is surrounded by a dotted line; a selected axis displays a selection handle at the center point of its range and is surrounded by a dotted line.
Figure 9-46 Dragging an Axis to Rescale a Ternary Plot Range
3. Drag either the apex or the axis toward or away from the center of the graph. The axis ranges adjust accordingly. Note: Modifying axis ranges of ternary graphs often introduces additional axes. These axes are the second axes of each "pair of axis lines. An axis which appears as a result of moving an apex is paired with the axis opposite the apex which moved. Additional axes can be modified and are controlled in the same way as the three original ternary axes using the Axes tab of the Graph Properties dialog box.
Figure 9-47 The left graph Y axis was dragged to 50%. The right graph Y apex was dragged to 50%.
482 Chapter 9
2. Click the Axes tab. 3. Select Scaling from the Settings for list. 4. Use the slider controls for X Range, Y Range, and Z Range to change individual axis ranges. Note: When you change the Minimum for any axis, the maximums for the other axes adjust automatically. The Maximum value must be greater than the Minimum value. 5. Click OK.
Note: Increasing an axis range minimum reduces the size of the ternary graph because it is always reduces the other axis range maximums. Reducing the maximum of a ternary axis range changes the graph shape.
6. Double-click the angular axis. The Graph Properties dialog box appears.
484 Chapter 9
Figure 9-50 Graph Properties Dialog Box Axes Tab Scaling Settings
7. Click the Axes tab. 8. Select Scaling from the Settings for list. 9. Select the new axis scale type from the Scale Type drop-down list. 10. Click OK. When you change the axis scale type for one axis, it is changed for all axes.
Ternary graph axes have interdependent axis ranges from 0 to 100, where 0 to 100 is the default setting or 0-1.0 where 0-1.0 is the default setting. For more information, see Ternary Scale Type above. The axis range and scale control the axis units and increments used to plot data.
To modify the axis direction:
2. Click the Axes tab. 3. Select Scaling from the Settings for list. 4. Select the axis you wish to modify from the Axis drop-down list. 5. Select the axis direction from the Direction drop-down list. 6. Click OK. The tick directions change on all three axes and the axis ranges reverse.
486 Chapter 9
Changing the axis directions inverts the 0-100 direction of the labels and changes the direction of the tick marks. However, axis titles only move if they are positioned along an axis, not at an apex. Apex position for each variable remain constant regardless of axis direction.
Figure 9-52 Ternary Graphs Displaying Counterclockwise (Left) and Clockwise (Right) Axis Directions
Figure 9-53 The X Axis ticks and labels are drawn in light gray, the Y Axis ticks and labels are drawn in black, and the Z Axis ticks and labels are drawn in dark gray.
1. Double-click the tick marks. The Graph Properties dialog box appears.
488 Chapter 9
Figure 9-54 Graph Properties Dialog Box Axes Tab Ticks Settings
2. Click the Axes tab. 3. Select Ticks from the Settings for list. 4. Select either Major Ticks or Minor Ticks from the Apply to drop-down list. 5. To turn tick drawing on and off and to select tick directions for both sides of an axis line, use the Direction lists. The second list is only available if a ternary plot range change has created a secondary axis. 6. Select Out, In, or In and Out to display tick marks on the selected axis out from the center of the graph, in toward the center of the graph, or both outward and inward. Select a clockwise, counterclockwise, both, or 90 Degree option to select the tick mark direction along the axis. Select (none) to hide tick marks.
Figure 9-55 Graph Examples of Tick Marks Pointing, counterclockwise, Clockwise, Both, and 90 Degrees
7. To change major tick intervals, move the Major Tick Intervals slider. 8. To change minor tick intervals, under Tick Intervals, select a new value from the Minor Tick Intervals drop-down list.
Figure 9-56 Ternary Graphs with Tick Intervals of 15 and 30
9. Click Apply. 10. Use the Axis drop-down list to modify tick marks on a different axis, or use the Apply to drop-down list to switch to modifying major or minor tick marks. 11. Click OK.
490 Chapter 9
Figure 9-57 Graph Properties Dialog Box Axes Tab Ticks Settings
2. Click the Axes tab. 3. Select Ticks from the Settings for list. 4. Select either Major Ticks or Minor Ticks from the Apply to drop-down list. 5. To change tick length and thickness, under Tick Line, move the Length and Thickness sliders. Drag the slider control with the mouse or set the tick length and thickness to specific values by typing directly in the Length and Thickness boxes. 6. To change tick color, under Tick Line, select a color from the Color drop-down list. Choose from any of the listed colors, or select (Custom) to use a pre-defined custom color or create your own color. Select (none) to create transparent tick marks. 7. Click Apply. 8. Use the Axis drop-down list to modify tick marks on a different axis, or use the Apply to drop-down list to switch to modifying major or minor tick marks.
9. Click OK.
1. Double-click the axis you want to change. The Graph Properties dialog box appears.
Figure 9-58 Graph Properties Dialog Box Axes Tab Labels Settings
2. Click the Axes tab. 3. Select Labels from the Settings for list.
492 Chapter 9
4. Select the Major (or Minor) Tick Labels check boxes. Depending on the selected axis, the check boxes are Top, Bottom, Left, or Right. 5. To change the direction of the axis tick labels, select the Clockwise and counterclockwise (CCW) check boxes. You can draw in both directions at once. 6. To draw tick labels at the 90 degrees tick position, clear both direction options.
Figure 9-59 Ternary Graph Axes with Tick Labels counterclockwise, Both Clockwise and CounterClockwise, and Neither (90 Degrees)
7. To add a suffix or prefix to the major or minor tick labels on ternary axes, select either Major Ticks or Minor Ticks from the Apply to drop-down list, then use the Add To Major (or Minor) Tick Labels options to type a prefix or suffix to the major or minor tick labels. 8. To rotate major or minor tick labels parallel to their axis, select either Major Ticks or Minor Ticks from the Apply to drop-down list, then under Rotate with Axis, select Tick Labels. 9. Click Apply. 10. Use the Axis list to modify tick labels on a different axis, or use the Apply To dropdown list to switch to modifying major or minor tick labels. 11. Click OK. Note: Tick labels and tick marks are controlled by their axis of origin, but may be drawn on axes other than their own.
Chapter
10
Use the Report Editor to annotate and document your graphs and data. The Report Editor features a complete text editor and OLE2 insertion and editing. It is also used by the Regression Wizard to report regression results. This chapter covers: Setting report options. For more information, see Setting Report Options below. Creating reports. For more information, see Creating Reports below. Exporting reports. For more information, see Exporting Reports below. Printing reports. For more information, see Printing Reports below. Using the Report Editor ruler. For more information, see Using the Report Editor Ruler below. Formatting text and paragraphs. For more information, see Formatting Text and Paragraphs below. Inserting the current date and time into a report. For more information, see Inserting the Current Date and Time into a Report below.
493
494 Chapter 10
Set whether or not you want to report only flagged values. Hide or display the report ruler.
To set report options:
The Options dialog box appears. 2. Click the Report tab. 3. To set the number of significant digits used for the values in the report, select Number of Significant Digits. The default is three digits. The maximum number of digits is sixteen. 4. To use scientific notation for the appropriate values in the report tables, select Always Use Scientific Notation. If this option is disabled, scientific notation is only used when the value is too long to fit in the table cell. This option is disabled by default. 5. To explain explanatory text for test results in the report, select Explain Test Results. This option is enabled by default. Clear this option to keep explanatory text out of the report. 6. To specify a significant P value, select P Value for Significance. This option determines whether there is a statistically significant difference in the mean values of the groups being tested. The value you specify is compared to the P values computed by all tests. Note: This P value does not affect the actual test results. It only affects the text that explains if the difference in the mean values of the groups is due to chance or due to random sampling variation. If the P computed by the test is smaller than the P set here, the text reads, "The difference in the mean values of the two groups is greater than would be expected by chance; there is a statistically significant difference between the input groups." If the P value computed by the test is greater than the P set here, the text reads, "The difference in the mean values of the two groups is not great enough to reject the
possibility that the difference is due to random sampling variability. There is not a statistically significant difference between the input groups." One of the above explanation text strings appears for each P value computed by the test. ANOVAs and some regressions produce multiple P values. Note: If the Explain Test Results option is cleared, the results of this P value do not appear in the report. 7. To display the ruler at the top margin of the report page, select Show Ruler. This option is enabled by default. Clear this option to hide the report ruler.
Creating Reports
Create reports using the New command, or the Regression Wizard.
To create a new report:
Right-click the section in the notebook where you want to create the report, and on the shortcut menu click New, and then click Report. A report window opens and a new report is added to the selected section.
496 Chapter 10
The page sample at the top of the dialog box reflects changes. 3. Select the paper size and source from the Size and Source drop-down lists. 4. To select the printer, click Printer. The Page Setup dialog box appears on which you can select and setup any printer configured for your system. 5. To change the paper orientation, under Orientation, select either Portrait or Landscape. 6. To change the margins, under Margins (inches), type the desired values into the four boxes. The current ruler units appear in the Margins title.
Exporting Reports
You can only export the entire report. If you want to export a portion of the report, delete the portion you do not want to export, then export the remainder as the file.
To export a report:
1. Select and view the report window you want to export. 2. From the menus select:
File Export
The Export File dialog box appears. 3. From the Files of type drop-down list, select a file format. 4. Enter the file name, directory, and drive for the exported file. 5. Click Export to create the file.
Printing Reports
You can print any report in a SigmaPlot notebook.
To display a report as it will look when printed:
1. Select and view the report window. 2. Click the Print button to print the report using all the default settings.
498 Chapter 10
The Print dialog box appears. 2. Click Properties. The printer Properties dialog box appears. 3. Click OK when you are satisfied with the printer properties settings. The Properties dialog box closes. Note: The Properties dialog box options vary from printer to printer. 4. Click OK to print the report.
Setting Tabs
All tab stops appear on the report ruler. The default tab stop is 0.25" regardless of the current units. Tab stops are made for individual and selected paragraphs, and are saved with reports.
To set a tab:
1. Select the paragraph(s) to change the tab stops. 2. Click the ruler where you want to place a tab. A tab marker appears at the clicked location. 3. To move a tab, drag the tab marker to another location on the ruler. To delete a tab, drag the tab marker off the ruler. You can also set tabs from the Tabs dialog box: 4. From the menus select:
Format Tabs
Note: This command is only available while viewing a report window. The Set Tab dialog box appears.
Figure 10-2 Set Tab Dialog Box
5. Enter tab stops in the Tab stop position (in) box. Enter Tab locations using the current ruler units. 6. Click OK to add the tab setting to the list.
500 Chapter 10
2. To change the first line indent, drag the marker at the top left of the ruler. 3. To change the left indent, drag the marker on the bottom left of the ruler. 4. To move both the left and first line indents, drag each marker separately. 5. To change the right indent, drag the marker on the bottom right side of the ruler. Note: To create an indented line, drag the top left marker to the right of the left indent. To create a hanging indent, drag top left marker to the left of the bottom left indent marker.
1. Select the text you want to modify. You can select individual characters, words, paragraphs, or the entire report. 2. To format character font, size, weight, underlining, or color, use the formatting toolbar buttons. For more information, see Formatting Text in Chapter 5. 3. To set paragraph alignment, use the Formatting Toolbar Align Left, Align Center, and Align Right and Justify.
502 Chapter 10
4. To add bullets or numbers to a selected paragraph, click the Bullet Style or Number Style button. To remove bullets, click the Bullet Style or Number Style button again. You can also right-click the report page and on the shortcut menu click Bullet or Number. Bullets are applied to the selected text.
3. Select the date and time format from the Available formats list. 4. Click OK. The current date and time appear as text at the specified location. Note: The list of available date and time formats depends on your Regional Settings. You can view or modify the Regional Settings directly from your Windows Control Panel.
Chapter
Publishing Graphs
11
You can use SigmaPlot 11.0 to publish graphs on the World Wide Web, and to create publication quality graphs for submission to journals and other printed forms. This chapter covers: Publishing graphs on the World Wide Web. For more information, see Publishing Graphs on the World Wide Web below. Submitting graphs for publication.For more information, see Submitting Graphs for Publication below.
503
504 Chapter 11
Using the SigmaPlot WebViewer, you can: View the graphs in high resolution. Pan and zoom the graph without losing resolution. Print in high resolution (printer resolution) as opposed to typical Web graphics (GIFs, JPEGs, etc.) that are printed in low resolution. View the data used to create the graph.
1. Open a graph page. 2. Select the page objects you want to publish. 3. On the File menu, click Save As Web Page. The Export File dialog box appears. 4. Enter a name of the file in the File name box. HTML (SigmaPlot Web Graph) already appears in the Save as type box. 5. Click Export. The Export Web Graph dialog box appears.
6. To set the size of the figure, select desired measurements from the Height and Width drop-down lists. Note: One inch is 96 pixels, and the Export Web Graph dialog box uses a fixed aspect ratio. 7. To export the currently selected graph or objects, select Export Selected Only. 8. To export the entire graph page, clear Export Selected Only. 9. To password protect the file, click Set Password. 10. To password protect the file, click Set Password. For more information, see Password Protecting Data on the World Wide Web below. 11. Click OK. Three files are created: an .HTM file which references a saved .JPG file and a .JNB file. You can later insert this .HTM file into any HTML editor.
506 Chapter 11
To set a password:
1. On the Export Web Graph dialog box, click Set Password. The Set Password dialog box appears.
Figure 11-2 Set Password Dialog Box
1. Select the graph on the page, and copy it. 2. On the Standard toolbar, click the New Page button. The Graph Page dialog box appears asking if you would like to create a graph. 3. Click No. 4. Paste the graph to the new page.
Now when you export this graph, you will only export the data associated with the graph.
1. Export a graph into HTML format. 2. In FrontPage, place the cursor on the page where you want to insert the WebViewer graph. On the Insert menu, click File. The Select File dialog box appears. 3. Select the HTML file you created in SigmaPlot to import into your FrontPage project, and click Open. A Javascript object containing WebViewer graph information appears at the insertion point on the page. The graph is not visible on the page until viewed using Internet Explorer.
508 Chapter 11
and other publications. These tend to vary, but are usually available at the web site for the journal submission requirements.
1. Determine the final size of the figure, the heights of text and thicknesses of lines and whether the figure will be color, grayscale, or black and white. 2. Determine what file formats are acceptable, and choose the best one. The ranking in which you should choose your format is: SigmaPlot EPS TIFF 3. Printed hardcopy (not really a file, but some publications actually still prefer this). These formats are regardless of whether the graph is color or not. Some publishers will directly accept SigmaPlot files. Most others accept EPS, TIFF, or both. 4. Determine how much the figure is going to be scaled using the size of your current figure. For example, if your graph is 5 inches wide, but the figures are printed at 3.25 inches wide, then scale your graph by a factor of 3.25/5, or .65.
5. Increase text labels and line widths accordingly on your SigmaPlot graph. For example, if you reduce your graph to .65 of the original size, and text must be 10pt in height, increase your labels to at least 15.5pt. Alternately, you can reduce the graph itself to the final publication size. 6. Make any other changes to your graph to meet the publishers requirements, such as typeface, labeling, and so on. 7. Once you have your graph formatted, produce the selected file. Make sure that you select the figure (click it) before choosing export-this will automatically crop your figure for you. If you are producing an EPS file, you do not need to pay attention to dpi at all. If you must use TIFF format, make sure you use the CMYK-compressed TIFF format. Uncompressed TIFF files are too big to easily handle. Also, you will now have to do some dpi calculations. For example, if you are producing a file that requires a final printed dpi of 600, and the graph is being reduced by a .65 ratio, do not set the file dpi to 600. Instead, use a dpi of 390 (600*.65). When this file shrinks to the final printed size, the final dpi will also be 600.
510 Chapter 11
About dpi
dpi (dots per inch) is a printer term, and is often misleading. dpi determines how many pixels are used to create the figure. A more accurate term would be resolution. You can increase the final dpi of a raster figure by shrinking it. This creates more pixels within a smaller space, increasing the dpi. Most printed figures do not require a dpi higher than 600 for grayscale figures, and 300 dpi for color figures. The 1200 dpi number is for black and white figures only that have no half toning. If you must produce a 1200 dpi figure, you will have to do some post-processing on your file in order make it palatable to the printer. This can be beneficial if you must use TIFF file and have Photoshop.
Use the Line and Text Properties dialog boxes to make global changes to your graphs before publishing.
or
Format Text Properties
Resizing Graphs
If you need to resize you graph for publication, set your fonts and line widths first, then turn the automatic re-scaling of these objects off before resizing your graph.
To resize your graph for publication:
The Options dialog box appears. 2. Click the Page tab. 3. Clear Graph objects resize with graph.
To re-scale the graph precisely:
512 Chapter 11
Select the graph before you export; otherwise, you will export the entire page including unnecessary white space surrounding the graph.
Disk Space and Memory
Make sure you have enough disk space and memory before trying to export a large graphic file. For a large file, you need to have at least 200 megabytes or more free on both your system drive (for swap and temp file space) as well as the same space on your destination drive. You can also increase your Virtual Memory to a very large size, but this is not necessary if you have sufficient hard drive space available. Note that it can take awhile to generate these files, depending on your systems speed and available RAM.
The Submission Assistant dialog box appears. 2. Pick a journal profile from the Submission Profile drop-down list. Note: Profile data is stored in individual profile files in your Users Folder. For more information, see About SigmaPlots User and Program Files in Chapter 1. 3. Enter the final figure size into the Height and Width fields. This is either set by default or you enter this information manually. 4. Select the file name and path to save the scaled figure. If you dont change the path name, the page name is used for the file name. 5. Select Save scaled figure to new page to review the scaled figure in SigmaPlot. 6. Clear Export selection only, if available, to export all graphs on the page. 7. Click Next. 8. The Submission Assistant - Summary dialog box appears with a list of Submission Criteria and a result of either Pass, Fail, or Warning in the Pass/Fail column. 9. Select any item in the Submission Criteria list. If the graph failed, the Details section offers suggestions on how to repair the graph to meet the submission standards. 10. Once the graph has met the Submission Criteria, click Export.
1. On the Submission Assistant - Edit profile dialog box, click Edit. The Submission Assistant - Edit profile dialog box appears. 2. Once you have edited the profile, click OK to save it.
514 Chapter 11
Chapter
12
SigmaPlot uses a VBA-like macro language to access automation internally. However, whether you have never programmed, or are an expert programmer, you can take advantage of this technology by using the Macro Recorder. This chapter describes how to use SigmaPlots Macro Recorder and integrated development environment (IDE). It also contains descriptions of related features accessible in the Macro window, including the Sax Basic programming language, debugging tool, dialog box editor, and user-defined functions. Record a macro any time that you find yourself regularly typing the same keystrokes, choosing the same commands, or going through the same sequence of operations.
515
516 Chapter 12
Recording Macros
To record a macro: 1. On the menus click:
Tools Macro Record New Macro
REC appears in the status area of SigmaPlots main window, indicating that the macro is recording your menu selections and keystrokes. 2. Complete the activity you want to include in this macro. Note: The Macro Recorder does not record cursor movements. 3. When you are finished recording the macro, on the menus click:
Tools Macro Stop Recording
The Macro Recorder stops recording and the Macro Options dialog box appears. 4. Type a name for the macro in the Name text box. Give the macro a descriptive name. You can use a combination of upper- and lowercase letters, numbers, and underscores. For example a macro that formats all of your graph legends to match a certain report might be called "Report1AddFormatToLegend". 5. Enter a more detailed description in the Description text box. 6. Click OK. After you have finished recording the macro, save it globally (for use in all of SigmaPlot) or locally (for use in a particular notebook file). Your macro appears in the Notebook Manager.
The New dialog box appears. For more information, see Creating New Items in the Notebook Manager in Chapter 2.
Figure 12-1 Select Macro in the New dialog box to create a new macro from scratch.
2. From the New drop-down list, select Macro. 3. Click OK. The Macro Window appears.
518 Chapter 12
Figure 12-2 A new Macro Window. You can create SigmaPlot macros from scratch using SigmaPlots VBAlike macro language.
Editing Macros
When you record a macro, SigmaPlot generates a series of program statements that are equivalent to the actions that you perform. These statements are in a form of SigmaPlot language that has custom extensions specifically for SigmaPlot automation and appear in the Macro Window. You can edit these statements to modify the actions of the macro. You can also add comments to describe code.
To edit a macro:
The Macros dialog box appears. 2. Select a macro from the Macro list. 3. Click Edit. The Macro Window appears. For more information, see Creating Macros Using the Macro Language above.
520 Chapter 12
Step Over. Executes to the next line. If the current line is a subroutine or a function call, execution of that subroutine or function call will complete. Step Out. Steps execution out of the current line the cursor is on. Step to Cursor. Steps execution out to the current subroutine or function call. Toggle Breakpoint. Toggles the breakpoint on the current line. The breakpoint stops program execution. Quick View. Shows the value of the expression under the cursor in the Immediate Window. Macros. Opens the Macros dialog box. Dialog Box Editor. Opens the Dialog Box Editor. Object Browser. Object Browser Reference. Editing Macros Opens the Reference dialog box which contains a list of all programs that are extensions of the SigmaPlot Basic language.
Color-Coded Display
The color-coding of text in the Macro Window indicates what type of code you are viewing. The following describes the default text colors used in the script text: Blue. Identifies reserved words in Visual Basic (for example, Sub End Sub, and Dim). Magenta. Identifies SigmaPlot macro commands and functions. Green. Identifies comments in your macro code. Separates program documentation from the code as you read through your macros.
The Options dialog box appears. 2. Click the Macros tab. 3. Set text colors for different types of macro code and Debug Window output. 4. Change font characteristics. 5. Set the location for the macro library.
522 Chapter 12
1. In the Macro Window, click where you want to edit. 2. You can also use arrows and key combinations to move the insertion point; when you do this the window scrolls automatically.
Open the macro code window and select the text to edit.
Adding Comments to Code
Add comments to code to identify the purpose of the various parts of a macro and to map locations as you edit a complex macro. Insert comments to fully document how to use and how to understand the macro code.
Deleting Unnecessary Code
The Macro Recordercreates code corresponding exactly to the actions that you make in SigmaPlot while the recorder was turned on. You may need to edit out unwanted steps.
When you need to find and change text in a macro that you have written, use the Find commands. For example, if you change the name of a file that is referenced in your macro, you need to change every instance of the file name in your macro. Use Find to locate the instances of the filename in the macro and replace using cut and paste edit commands.
1. In the Macro Window, place the insertion point where you want to put the code for the dialog box. For more information, see Editing Macros above. 2. On the Macro Window toolbar click the User Dialog button. The blank grid in the UserDialog Editor appears. 3. On the left hand side of the UserDialog Editor there is a Toolbox. You can select a tool, such as a button or check boxes, from the Toolbox. The cursor changes to a cross when you move it over the grid.
524 Chapter 12
4. To place a tool on the dialog box, click a position on the grid. A default tool will be added to the dialog grid. 5. Resize the dialog box by dragging the handles on the sides and the corners. 6. Right-click any of the controls that you have placed on the dialog surface (after selecting the control) and enter a name for the control. 7. Right-click the dialog form (with no control selected) and enter a name for the dialog monitor function in the DialogFunc field. 8. To finish, click OK. The code for the dialog box with controls will be written to the Macro Window. Finally, and in most cases, you must edit the code for dialog box monitor function to define the specific behavior of the elements in your dialog box.
1. On the Macro Window toolbar, click the Object Browser button. 2. Use Paste to insert generic code based on your selection into a macro. Tip: Press F1 at any time for full details on using the Object Browser.
statements, aside from calling subroutines that do the work. You add procedures using the Add Procedure dialog box.
To add a procedure:
1. On the Macro Window toolbar, click the New Procedure button. The Add Procedure dialog box appears. 2. Define a sub, function, or property using the Name, Type, and Scope boxes. 3. Click OK to paste the code for a new procedure. The new procedure appears at the bottom of the macro. Tip: For full details on using the Add Procedure dialog box, press F1 from anywhere in the Macro Window.
526 Chapter 12
Press F1 from anywhere in the Macro window to view user-defined function on-line Help.
Immediate Tab
The Immediate Tab lets you evaluate an expression, assign a specific value to a variable or call a subroutine and evaluate the results. Trace mode prints the code in the tab when the macro is running. Type "?expr" and press Enter to show the value of "expr". Type "var = expr" and press Enter to change the value of "var". Type "set var = expr" and press Enter to change the reference of "var" for object vars. Type "subname args" and press Enter to call a subroutine or built-in expression "subname" with arguments "args". Type "trace" and press Enter to toggle trace mode. Trace mode prints each statement in the Immediate Tab when a macro is running.
Watch Tab
The Watch Tab lists variables, functions, and expressions that are calculated during execution of the program. Each time program execution pauses, the value of each line in the window is updated. The expression to the left of the "->" may be edited. Pressing Enter updates all the values immediately. Pressing Ctrl+Y deletes the line.
528 Chapter 12
Stack Tab
The output from the Stack Tab lists the program lines that called the current statement. This is a macro command audit and is helpful to determine the order of statements in you program. The first line is the current statement. The second line is the one that called the first, and so on. Clicking a line brings that macro into a sheet and highlights the line in the edit window.
Sub SigmaPlot() SigmaPlot Objects and Collections SigmaPlot Macro Dim SPApp as Object Set SPApp = CreateObject("SigmaPlot.Application.1") SPApp.Visible = True SPApp.Application.Notebooks.Add End Sub 4.
4. To run the macro, from the menus select:
Run Run Sub/User Form
530 Chapter 12
The Macros dialog box appears. 2. Select a macro from the Macro Name scroll-down list.
3. Click Options. The Macro Options dialog box appears. 4. Select Command Name. 5. Enter the name of the macro in the Command Name field. If the Command Name is cleared, the macro does not appear on a menu. 6. Enter the name of the menu under which you want the macro to appear in the Menu Name field. 7. Click OK. Your new macro appears under the menu command you have just created. 8. Enter the same menu command name in the Menu Name field of future macros if you want them to appear on your new macro command menu. By default, if the Menu Name field is left empty, the macro name appears on the Tools menu. You can also create your own menu by entering the menu name in the Menu Name field.
532 Chapter 12
The Macros dialog box appears with a list of available macros. 2. Select the macro to run from the Macro name list. 3. Click Run.
SigmaPlots Macros
SigmaPlots available macros are: ROC Curve Analysis.For more information, see ROC Curve Analysis below. Area Below Curves. Integrates under curves using the trapezoidal rule. For more information, see Area Below Curves below. Batch Process Excel Files. Imports data from multiple Excel Files into individual SigmaPlot worksheets, then plots and curve fits the imported data automatically.For more information, see Batch Process Excel Files below. Border Plots. Draws a histogram or box plot along the top and right axes of a scatter plot. For more information, see Border Plots below. By Group Data Split. Splits data contained in one column into groups of data sorted into multiple data columns within one SigmaPlot worksheet. For more information, see By Group Data Split below. Color Transition Values. Creates a column of colors changing smoothly in intensity as the data changes from its minimum value to its maximum value. For more information, see Color Transition Values below. Compute 1st Derivative. Computes a numerical first derivative of a pair of data columns. For more information, see Compute 1st Derivative below. Frequency Plot. Creates frequency plots with mean bars for multiple data columns. For more information, see Frequency Plot below.
Gaussian Cumulative Distribution. Returns the results of a Gaussian Cumulative Distribution function (CDF) for a single column of data, and optionally plots the results with a probability Y axis scale. For more information, see Gaussian Cumulative Distribution below. Insert Graphs into Microsoft Word. Inserts a SigmaPlot graph into an open Microsoft Word document. For more information, see Insert Graphs into Microsoft Word below. Label Symbols. Labels a plot with symbols or text from a specified column. For more information, see Label Symbols below. Merge Columns. Merges two separate worksheet columns into one single text column. For more information, see Merge Columns below. Paste to PowerPoint Slide. Creates PowerPoint slides from selected SigmaPlot graphs. For more information, see Paste to PowerPoint Slide below. Piper Plots. Creates a Piper Plot. For more information, see Piper Plots below. Plotting Polar and Parametric Equations. Creates curves in either Cartesian or polar coordinate systems. For more information, see Plotting Polar and Parametric Equations below. Power Spectral Density. Computes the power spectral density (psd) for a data column. For more information, see Power Spectral Density below. Quick Re-Plot. Re-assigns the columns that are plotted for the current curve in the current two- or three-dimensional plot. For more information, see Quick Re-Plot below. Rank and Percentile. Computes ranks and cumulative percentages for a specified data column. For more information, see Rank and Percentile below. Survival Curve. Computes and graphs a Kaplan-Meier survival curve using the SurvlMod transform. For more information, see Survival Curve below. Vector Plot. Uses the vector transform to plot X,Y, angle and magnitude data as vectors with arrowheads.
534 Chapter 12
or {yi(xi+1 - xi) + (1/2)(yi+1 - yi)(xi+1 - xi)} Specify the column number for the results. Click Compute to calculate the area under the curve. The results are in the Results Column.
Restrictions
This macro only works using a SigmaPlot worksheet. This macro only works with plots with both X and Y data. A graph window containing a scatter or line plot must be open and in focus when running the macro.
Save notebook to. Shows the path where the BatchFile.jnb notebook is saved. Browse. Select to save the BatchFile.jnb notebook file to a folder other than the default. You may edit the macro to change: To a different fit library. The default location of the source data block. Whether the data are plotted or fit. The file extension to import different file types.
Restrictions
Only data from the first Excel worksheet from each file is imported. You cannot specify a different Excel worksheet. You must select an Excel file. You must select a curve in order to plot curve fit. You may change the default equation Only simple scatter and bar charts are available.
Border Plots
This macro draws a histogram or box plot along the top and right axes of a scatter plot. The border plots are located .5 inches from each axis. When using histogram border plots, specify the number of bins displayed.
Restrictions
A graph window containing a scatter plot must be open and in focus when running the macro. If the current plot is not a scatter plot, the macro can convert the plot to the required form. If the plot is an X only or Y only plot, the macro creates one border plot corresponding to the X or Y axis.
536 Chapter 12
To run the macro, a worksheet must be open and in focus. The worksheet must have at least two columns of data. Can accept empty cells. Puts "--" into grouped cell to indicate an empty cell or missing data. Group Column should not have empty cells. Grouped Data Column Heading Titles correspond to the Group Column contents. Output column entry must be numeric and greater than the last data column.
Restrictions
Frequency Plot
Creates frequency plots with mean bars for multiple data columns. The following options are available: Column Selection. Specify the number of data columns. The analysis includes all columns between the first and the specified column, inclusive. Each column corresponds to a group. In addition, define the column in which to begin placing the macro results. Graph Dimensions. Set the height and width of the frequency plot in inches.
538 Chapter 12
Bins. The Vertical Interval indicates the range into which data points will be grouped. The Start Value defines the smallest vertical interval value. A value of 0 corresponds to a vertical range from 0 to 100. Mean/Median Lines. Add a line (of the specified width) corresponding to the mean or median for each group. Symbols Size defines the diameter of the plotting symbols. Gap represents the horizontal distance between symbols as a percentage of the symbol diameter.
Restrictions
To run the macro, a worksheet must be open and in focus. Data must begin in column 1.
Restrictions
You must have both a Word Document and a SigmaPlot notebook open in order to use this macro.
Label Symbols
This macro labels a scatter or simple bar plot with text from a specified column. Select an offset percentage and one of eight locations for the labels. The offset percentage is a percent of each axis range - 3 to 5 percent is a good starting value. If you have more than one plot in your graph then select the plot you want to label by clicking on a symbol or bar of the particular plot. Rerun the macro to obtain the best position for the labels and then use Graph, Delete Plot to remove the unwanted label plots. You can label the plot with numbers by placing the numbers in a column and using Format, Cells to change the numbers to text. Then format each text-number for a pleasing appearance by removing places to the right of the decimal point, and so on. If the label column contains fewer entries than the plot contains symbols or bars, labeling continues by returning to the first case of the label column. For example, in a scatterplot containing six points, if the label column contains three entries (A,B,C), the points are labeled (A,B,C,A,B,C).
Restrictions
A graph window must be open and in focus when running the macro. The macro is restricted to scatter and simple bar charts and is not applicable to stacked or grouped bar charts.
Merge Columns
This macro merges two separate worksheet columns into one single text column. This is useful if you have two text fields that need to be combined into one, or if you have imported data that contains dates in one column and time in another. To run the macro, simply select the first and second columns to merge, then click OK.
540 Chapter 12
Restrictions
At least two columns of data must present on your worksheet. The results are automatically placed into the first empty column after the last data column.
1. Select New Presentation or Existing Presentation. If you select Existing Presentation, the Select Presentation button becomes active. 2. Click Select Presentation. The Select Presentationdialog box appears. 3. Navigate to your existing presentation, and double-click the presentation. The macro returns to the Insert Graph into PowerPointdialog box. 4. Select the PowerPoint Slide Layout you want to insert the graph into from the Slide Layoutdrop-down list. 5. Select the PowerPoint Slide Background you want from the Slide Backgrounddropdown list. 6. Select the SigmaPlot Graph Scheme from the drop-down list. 7. Select the Transparent Graph Plane check box if you want a transparent background for the graph. 8. Click OK. PowerPoint opens, and the graph object appears centered in a new PowerPoint slide.
To edit the graph object in PowerPoint:
1. Open SigmaPlot.
2. Open the PowerPoint presentation with the inserted SigmaPlot graph. 3. Double-click the graph in PowerPoint. The graph color changes, and the handles around the graph turn black. 4. Right-click over the graph. The SigmaPlot right-click menu appears. 5. Select Object Properties to edit graph Lines, and Size and Position. 6. Select Graph Properties to edit graph Plots, Axes, Grids and Planes, and Title and Legend. 7. Click OK to apply changes.
Piper Plots
PIPER plot showing the variations in groundwater chemistry in different sandstone aquifers of England courtesy Dr Paul Shand, British Geological Survey http://www.bgs.ac.uk/hydro/index.htm
Using the Piper Plot Macro
To create a Piper plot: 1. Make sure the values of your four cations and three anions are entered into seven columns. 2. Pick the units the data use. If the units used do not match any of the options (percentages, mg/l, or mmol/l) then you will need to transform your data to one of these units. 3. Assign the column for each cation/anion as desired by selecting a worksheet column from the Worksheet columns list, the cation/anion from the Assign to list, and clicking the Assign button. You column assignments are listed under Assigned columns. 4. When finished, click OK.
542 Chapter 12
The PIPER diagram (Piper, 1944) is a plot commonly used by hydrogeologists and hydrologists to display water chemistry data. It has the advantage that many different water samples can be plotted on one graph. The major dissolved ionic species in most natural waters are Ca2+, Mg2+, Na+, K+, Cl-, HCO3- and SO42-. The PIPER diagram displays the relative proportions of the major cations (positively charged ions) and anions (negatively charged ions) on two adjacent triangular plots (Figure 1). For plotting purposes, Na and K are grouped together. Each apex of the triangle represents 100% of that component and mixtures of components plot either along the axes (for 2 components) or within the triangle (for 3 components).
Figure 12-3 PIPER diagram showing how the relative proportions of cations and anions are plotted.
The PIPER diagram above shows how the relative proportions of cations and anions are plotted. The diamond shaped graph is used to represent the composition of water
with respect to both cations and anions. This graph has the advantage that mixing between two waters plots as a straight line. The cation and anion points for each sample are projected onto the diamond shaped field along a line parallel to the outer axes of each triangular plot as shown on Figure 1 and the intersection of these points is plotted. The PIPER diagram can be used to classify "hydrochemical facies" or "water-types" based on the dominant ions. The proportions of different elements provide information on the chemical history of groundwaters and indicate the dominant chemical reactions that occur between groundwater and the rocks through which it passes. Some data from a sandstone aquifer in Yorkshire are shown on Figure 3 where it can be seen that the groundwaters generally vary from a Ca-HCO3 type to a Ca SO4 type due to the dissolution of the minerals calcite (CaCO3) and gypsum (CaSO4). Some samples also trend towards the Na and Cl apex of the plots due to mixing with an old seawater component (dominated by Na and Cl). Piper, A.M. 1944 A graphic procedure in the geochemical interpretation of water analysis. Transactions, American Geophysical Union, 25, 914-923.
Figure 12-4 The distribution of water types from a sandstone aquifer in the Vale of York displayed on a PIPER diagram.
544 Chapter 12
The PIPER diagram above displays the distribution of water types from a sandstone aquifer in the Vale of York.
Quick Re-Plot
This macro quickly re-assigns the columns that are plotted for the current curve in the current two- or three-dimensional plot. Click Next Curve to change the plotted columns for several curves simultaneously.
Restrictions
546 Chapter 12
Restrictions
optimal cutoff point on the ROC curve will move from the vicinity of the upper left corner over toward the upper right corner. In prostate cancer screening, however, because benign enlargement of the prostate can lead to abnormal (high) PSA values, false positives are common and undesirable (expensive biopsy, emotional impact). In this case maximizing specificity is important (moving toward the lower left corner of the ROC curve).
An important measure of the accuracy of the clinical test is the area under the ROC curve. If this area is equal to 1.0 then the ROC curve consists of two straight lines, one vertical from 0,0 to 0,1 and the next horizontal from 0,1 to 1,1. This test is 100% accurate because both the sensitivity and specificity are 1.0 so there are no false positives and no false negatives. On the other hand a test that cannot discriminate between normal and abnormal corresponds to an ROC curve that is the diagonal line
548 Chapter 12
from 0,0 to 1,1. The ROC area for this line is 0.5. ROC curve areas are typically between 0.5 and 1.0 like shown in above. Two or more tests can be compared by statistically comparing the ROC areas for each test. The tests may be correlated because they occurred from multiple measurements on the same individual. Or they may not be correlated because they resulted from measurements on different individuals. The ROC Curves Analysis Module refers to this as "Paired" and "Unpaired", respectively, and can analyze either situation. The test measurements may contain missing values and two methods are provided to handle missing values when comparing ROC areas- pair-wise deletion and casewise deletion. This is described in detail later. Given a value for the probability that the patient has the disease (pre-test probability) the probability that the patient has the disease, given the value of the test measurement, can be computed. Also, given a value for the false-positive/falsenegative cost ratio (for the screening example above, the false-negative cost would be greater than the false-positive cost), an optimal test value cutoff can be computed. The present module allows entry of the pre-test probability and the false-positive/falsenegative cost ratio.
Data Entry
You can enter data in two formats: Indexed Grouped Indexed Data Format This is the format found in statistics programs such as Systat and SigmaStat. "Indexed" is the terminology used in SigmaStat. It has one column that indexes another column (or other columns). It is also the format of the output of logistic regression where ROC curves are used to determine the ability of different logistic models to discriminate negative from positive test results (normals from abnormals). Each data set consists of a pair of columns a classification variable and a test variable. The classification variable has a binary state that is either negative (normal) or positive (abnormal). Many programs use a value of 1 for positive and 0 for negative. The classification variable is required to be located in column 1 of the worksheet. The test variable is a continuous numeric variable and contains the test results. A single test variable will be located in column 2. Multiple test variables will be located in multiple columns starting in column 2. There is no built-in limit for the number of test variables. There is only one
classification variable for multiple test variables and it is located in column 1. The test variable columns must be left justified and contiguous. Therefore, no empty columns to the left of or within the data are allowed. The following example shows a few rows of data for two data sets. The first column is the classification variable. It contains a column title "Thyroid Function", which is the classification variable name. It also contains the two classification states "Hypothyroid" and "Euthyroid" (normal thyroid function). Hypothyroid and Euthyroid are the abnormal and normal classification states, respectively. T4 and T5 are the names of different blood tests that will be used in the ROC analysis to discriminate between normal and abnormal and then compared to determine which is the better test. The classification variable must be in column 1 and the two test variables in the two columns adjacent to it. The classification variable name will be obtained from the column 1 column title if it exists. The test names will be obtained from the column titles of the test variable columns if they exist. The classification state names will be obtained from the entries in the cells of column 1. If no column titles have been entered for the test variables then default names for the tests, "Test 1", "Test 2", etc., will be used and displayed in the graphs and reports. The test variable names should be unique but the program will subscript any identical names that are not.
Figure 12-5 Indexed data format for two tests.The test names are T4 and T5, the classification states are Euthyroid and Hypothyroid and the Classification variable name is Thyroid Function. The index column is always column 1 and data columns must be left adjusted.
There must be two or more non-missing data points for each test for each classification state. Missing values are handled automatically by the analysis. For data columns,
550 Chapter 12
missing values are everything but numeric values (blank cells, the SigmaPlot doubledash missing value symbol, "+inf", "-inf", "NaN", etc.). Missing values are ignored for all computations except the Paired area comparison (see the Missing Value Method section) where they are handled using one of two possible algorithms. Grouped Data Format The grouped data format consists of pairs of data columns - one pair for each test. One column in a data pair consists of the negative (normal) data values and the other column for positive (abnormal) values. So, for example, if two tests are to be compared, the worksheet will contain four columns of data - the first two columns for the first test and the third and fourth column for the second test. A specific column title format is used to identify the test associated with the data column pair and the classification states within each pair. The user is encouraged to use this format since it clearly identifies the data in the data worksheet and will annotate all the graphs and reports generated. It is not necessary to use column titles as the program will identify column pairs starting in column 1 with the generated test names "Test 1", "Test 2", etc., and will arbitrarily assign "1" and "0" classification state names to the first and second columns, respectively, but this is clearly not the best way to organize the data. Since the test names and classification states are numerical it is also more difficult to interpret the results. Column Title Convention for Grouped Data This column title convention is a simple way to identify worksheet data for the Grouped data format. The following example shows a few rows for two data sets. The first two columns contain the data for the T4 test. The first column T4 Euthyroid is the column with the normal data for test T4. The column title consists of the test name followed by a minus sign followed by the classification state. Spaces on either side of the minus sign are ignored. The second column T4 Hypothyroid is the column with the abnormal data for test T4. The third and fourth column titles are the same as the first two except the second test name T5 is used.
Figure 12-6 Grouped data format for two tests. This is the same data as in the figure above. There are two tests T4 and T5. Each test consists of a pair of data columns. In this case T4 is in columns 1 and 2 and T5 in columns 3 and 4. The Test-State column title format is used to identify the two tests and the normal (Euthyroid) and abnormal (Hypothyroid) states.
The test names in both columns of a column pair must be the same. Also there must be exactly two classification states in the column titles. Like Indexed format, missing values in the worksheet cells are ignored except for special handling when comparing ROC areas (see the Missing Value Method section).
The ROC Curve Analysis dialog box appears. Data Format. In most case the program will identify the data format from the information in the data worksheet. In the dialog above the format was identified as Indexed. You may select from the two formats: Indexed and Grouped. Data Type. If two or more data sets are selected then the Data Type option for correlated tests is available. You may select either Paired, for correlated tests, or Unpaired. If Paired is selected the ROC areas and area comparisons are determined using the DeLong, Delong and Clarke-Pearson method(2). If Unpaired is selected the areas are computed using the Hanley and McNeil method(3) and the areas are compared using a Z test.
552 Chapter 12
Missing Value Method. If missing values exist then two options are available for the pairwise comparison of ROC areas - Pairwise Deletion and Casewise Deletion. This option is not available if no missing values exist. Pairwise deletion only deletes rows containing missing values for the particular pair being analyzed not for an entire row of data. Fewer data values are deleted using this method. There are situations when pairwise deletion will fail but this is the option to use when it is possible. Casewise deletion deletes all cells in any row of data containing a missing value. Much more data may be deleted using this option. To better understand the difference, consider a simple example of two data columns of equal length one of which has no missing values and the other has one missing value. When ROC areas are being compared, certain computations on these two columns will be done pairwise - the first column with itself, the first column with the second column and the second column with itself. When the column without a missing value is being compared with itself no row deletions occur for pairwise deletion. For casewise deletion, however, the row that contains the missing value will be deleted from both data sets. So, for casewise deletion, the computation involving the column without a missing value with itself will be done with one row deleted (the row corresponding to the missing value in the other data set). The module determines when pairwise deletion is not valid and informs the user when this is the case. Positive State Options - Classification State and Direction. The two classification states are referred to as "Negative" (normal) or "Positive" (abnormal). The ROC analysis software must be informed which state is "Positive" and whether the test measurement values for the positive state are "High", meaning higher than those of the negative state, or "Low", meaning lower than those of the negative state. Accepted normal values for the PSA (prostate specific antigen) test are less than 4 ng/ml and abnormal values are higher than this. Thus if the two classification states names are "positive" and negative then the Positive state is "positive" and the Positive Direction is "High". In this case you would select the radio button next to "positive" and "High". On the other hand, for the T4 (thyroxine) test for hypothyroidism the T4 values are lower in the abnormal state than for the normal state. In this case the abnormal Positive State is "Hypothyroid" and the Positive Direction is "Low". So you would select the radio button next to "Hypothyroid" and "Low". What happens if you select the incorrect option? Sensitivity (specificity) is defined in terms of the positive (negative) state. So if the positive state is incorrectly selected then sensitivity and specificity will be incorrectly defined (switched) and the ROC curve will have the X and Y axes switched. This will result in an ROC curve that
appears below the diagonal unity line. It will have an area less than 0.5. The module will detect this and give you the options. It is possible that there is something wrong with the data so you can Abort the analysis and correct the problem. More likely you have selected the incorrect positive state or direction so you can Retry the analysis with correct selections. In rare occasions for multiple tests some tests will have areas greater than 0.5 and one or more will have areas less than 0.5. In this case you can Ignore this warning and continue with the analysis. Available Data Sets - Selected Data Sets. Select one or more of the available data sets by clicking on them in the Available Data Sets window and then clicking on the Add button. If desired, you may then select a test name in the Selected Data Sets window and click Remove to deselect the test. Reports Confidence Intervals. Confidence intervals are computed for statistics in both the Sensitivity and Specificity and Area Comparison reports. You can generate 90, 95 and 99% confidence intervals. Create Sensitivity and Specificity Report. Cutoff values are created between each test data value in the (sorted) data set. If there are a large number of data points and several tests then there will be a large number of cutoff values and the Sensitivity & Specificity Report can be very long. If you clear this option then all report options in the dialog below this are not required and are disabled. Fractions/Percents. You may display sensitivities, specificities and probabilities in either fraction or percent format. Selecting Percents also requires the pre-test probability to be entered as a percent. Create Post-Test Results. Selecting this option allows entry of the pre-test probability. It also enables the possible entry of the false-positive/false-negative cost ratio. Given a pre-test probability the program will create post-test probabilities, both the positive predictive value (PV + = probability of disease given a positive test result) and the negative predictive value (PV - = probability of no disease given a negative test result), for each cutoff value. If the cost ratio option is selected then the optimal cutoff value will be computed. All of these results are displayed for each test in the Sensitivity & Specificity report. ROC Graph Options. All of the graph options in the dialog apply to the ROC graph. They allow you to add a diagonal line to the graph, add grid lines, add symbols for sensitivity and specificity at each cutoff point and change the ROC plot lines from solid to different line styles.
554 Chapter 12
The ROC curves graph for three data sets is shown below. These graphs are derived from numerical results in the worksheet entitled Graph Data. The graph title is obtained from the section name containing the raw data. The legend shows the test names and the ROC areas for each curve. The diagonal line and grids options were selected for this graph.
Of course, you can edit this graph in any way you wish. You might want to change the starting color of the color scheme used for the line colors. You can do this by doubleclicking on one of the ROC plot lines and then right-clicking on the Line Color dropdown list as shown below.
Figure 12-9 Selecting line colors for an ROC Curves Graph.
Dot Histogram Graph
556 Chapter 12
Dot histograms for the data associated with the ROC curves are shown below.
Figure 12-10 Dot histogram pairs for each test. The horizontal lines and the tables below the graph show the optimal cutoff values determined from the pre-test probability and cost ratio.
The graph title is obtained from the title of the section containing the raw data. The xaxis tick labels are obtained from the test names and the classification state names. The tick labels will rotate if they are too long to fit horizontally. The symbol layout design allows for symbols to touch horizontally and nest vertically. If values for pre-test probability and false-positive/false-negative cost ratio are entered then the optimal cutoff values for each test are computed and represented as a horizontal line across the two dot histograms for each test. The numeric values for the
optimal cutoff parameters are shown as tables below the x-axis. Sensitivity and Specificity Report The sensitivity and specificity report contains results for all tests with additional tests results placed in report rows below those of prior tests. The results for each test can be separated into three parts: Optimal cutoff value. Sensitivity and specificity versus cutoff values. Likelihood ratios and post-test probabilities. If values for both pre-test probability and cost ratio have been entered then the optimal cutoff is calculated. A slope of the tangent to the ROC curve m is defined in terms of the two entered values (P = pre-test probability). The optimal cutoff value is computed from sensitivity and specificity using the slope m by finding the cutoff that maximizes the function The results of this computation in the Sensitivity & Specificity report are shown below.
Figure 12-11 Optimal cutoff results in the Sensitivity and Specificity report.
For this data set, the optimal cutoff is 7.125 for a pre-test probability of 0.5 and cost ratio of 1.0. Sensitivities, specificities and their confidence intervals are listed as a function of cutoff value in the second part of the report. A portion of these results is shown in the table below. These results can be expressed as fractions or percents by using the Fractions/Percents option.
558 Chapter 12
Figure 12-12 Sensitivity and specificity results in the Sensitivity and Specificity report.
The third part of the Sensitivity and Specificity report contains the likelihood ratios and post-test probabilities.
A portion of the report showing the likelihood and post-test probabilities results is shown below.
Figure 12-13 Positive and negative likelihood ratios, LR+ and LR-, and post-test probabilities, PV+ and PV-, in the Sensitivity and Specificity report.
The positive likelihood ratio is not defined for some cutoff values since specificity = 1. ROC Areas Report The ROC Area report consists of two parts: ROC areas and their associated statistics. Pairwise comparison of ROC areas. An example of a report is shown in below.
560 Chapter 12
Figure 12-14 An example ROC Areas report. From top to bottom it shows the type of analysis used together with the missing value method, the ROC areas and associated statistics and a pairwise comparison of ROC areas.
In this case there are three correlated tests. Row two of the report shows that a Paired Analysis was performed and, since there were missing values in the data, Pairwise Deletion of missing values was selected to compare the areas. The first section of the report shows the ROC curve areas for the three tests. This is followed by the standard error of the area estimate, the 95% confidence interval (90% and 99% are also available) and the P value that determines if the area value is significantly different from 0.5. The sample size and the number of missing values for each classification state are given. The number of missing values reflects only what is seen in the data and does not give the number used for each computation-pair in the pairwise-deleted comparison of areas. The second section shows the results of the pairwise comparison of areas. The method of DeLong, DeLong and Clarke-Pearson(2) is used to compare areas when the Paired data type option is selected. When the Unpaired data type is selected, areas are compared using a Z test. The report shows results for all pairs of data sets. The difference of each area pair and its standard error and 95% confidence interval are computed. This is followed by the chi-square statistic for the area comparison (or Z
statistic if Unpaired is selected) and its associated P value. Formatted Full Precision Display This report presents the numeric results in a four significant digit format with full precision available. Double click on any cell (except the confidence intervals) to display the number at full precision. Additional Graphs Results data in both reports can be used to create additional graphs. Some examples seen in the literature are shown here. Sensitivity and Specificity versus Cutoff The data for the graph in the graph below is from the Sensitivity & Specificity report in columns 1, 2 and 4. Use the Data Sampling option in Graph Properties, Plots, Data to specify the row range for the graph (you can also drag select the rows in the worksheet to do this).
Figure 12-15 Graph of sensitivity and specificity versus. cutoff for one test using data from columns 1,2 and 4 of the Sensitivity & Specificity report.
Likelihood Ratios
562 Chapter 12
The positive and negative likelihood ratios for three different imaging modalities are shown in below (the data is artificial). The data is in columns 1, 6 and 7 of the Sensitivity & Specificity report. The values associated with the optimal cutoff are shown as solid symbols. The largest positive likelihood and smallest negative likelihood at the optimal cutoff is associated with magnetic resonance imaging (MR).
Figure 12-16 Positive and negative likelihood ratios graphed from data in the Sensitivity and Specificity report from columns 1, 6 and 7. The results for three tests are shown together with values associated with the optimal cutoff (solid symbols).
Optimal Cutoff versus Cost Ratio
Frequently it can be difficult to determine a value for the false-positive/false-negative cost ratio. So it is worth performing a sensitivity analysis (sensitivity here means how much one variable changes with changes in a second variable) to see whether the cutoff value changes significantly in the range of cost-ratio values of interest. The ROC Curves Module was run multiple times for different cost ratios and a graph of optimal cutoff vs. cost ratio for the three imaging modality tests is shown below.
Figure 12-17 Optimal cutoff values obtained from multiple runs of the module. Regions of insensitivity, or strong sensitivity, to cost ratio can be identified.
If the relative cost of a false-positive is much greater than that of a false-negative then the cost ratio is greater than 1. But lets assume that we dont know exactly how much greater it is but have some idea that it should be in the range of 2 to 5, say. Looking at the optimal cutoff for the best imaging modality (MR, green line) we find that it doesnt change for cost ratios from 2 to 20. So the optimal cutoff is insensitive to cost ratio and, in this case, it is not important to know a precise value for cost-ratio. Post-Test Probability Versus Pre-Test Probability Given values of sensitivity and specificity associated with the optimal cutoff a graph of post-test probabilities as a function of pre-test probability can be created using equations (5) and (6). The post-test probability of disease when the test is positive, blue lines in the graph below, was obtained from equation (5) and the post-test probability of disease when the test was negative, red lines, was obtained from 1.0 minus equation (6). A transform was written in SigmaPlot implementing these two equations that
564 Chapter 12
generated the post-test probabilities for a range of pre-test probabilities. The results for the best test, MR, and worst test, US, are shown. The MR test is clearly better since the post-test probability range, from negative test to positive test, is larger. Thus given a positive test the patient is more likely to have the disease using the MR test rather than the US test. Similarly, given a negative test it is less likely that the patient has the disease using the MR test.
Figure 12-18 Post-test probabilities of disease given positive and negative test results. The MR test is based on sensitivity = 0.94 and specificity = 0.97 whereas the US test used sensitivity = 0.78 and specificity = 0.85.
Standard Curve
A standard curve is used to calibrate an instrument or assay. The Standard Curves macro provides five equations that may be fit to your data. These range from a straight line equation to two different five parameter logistic equations.
The X data may or may not be logarithmic and, if not, may still be graphed logarithmically. Multiple Y replicate columns may be used. Predicted values may be obtained after the curve fit is performed: Y values from Xs, X values from Ys, and ECxx values from xx percentages. These predicted values may be added to your graph as symbols with drop-lines to the X and Y axes. The Dynamic Curve Fitting algorithm may be used to help solve difficult curve fitting problems involving local minima.These are typically encountered with the five parameter logistic functions.
566 Chapter 12
values for a specified range of percent values. You can also elect to plot the results of these on your standard curve. 8. When finished, click OK. A standard curve appears, and if you elected to compute additional values, they are also plotted using drop lines to indicate the X and Y values.
Figure 12-19 A Standard Curves Graph
Y Replicates
If you have multiple measurements for each X value, select Y replicates on the Standard Curve dialog box, then select the Last Y replicate column from the dropdown list. The Y replicate columns must begin to the right of the X data.
Linear Equation
y = y0 + ax
A straight line, characterized by the slope a and the y-intercept y0.
Quadratic Equation
y = y0 + ax + bx2
The standard parabolic equation with quadratic coefficient b, slope a and intercept y0.
This is a typical dose-response curve with a variable slope parameter. Four parameters are produced: Min, or bottom of the curve. Max, or top of the curve. EC50: Median effective concentration. That is, the concentration that can be expected to cause a defined effect on 50% of a given population of organisms under defined conditions.
568 Chapter 12
where
Richards formulation of the five parameter logistic adds an asymmetry parameter s to the four parameter logistic. The asymmetry is shown above with large changes in curvature with changes in s in the lower curve but relatively small changes in the upper curve. The additional algebraic equation for xb maintains EC50 as the half-maximum y value. The equation has been written so that a positive Hillslope results in a curve that increases with x. Four of the five parameters are the same as those in the four parameter logistic. s controls the asymmetry.If s = 1 then this function is the same as the four parameter logistic.s less than 1 decreases the overall slope of the curve whereas s greater than 1 increases the overall slope.
where
Figure 12-21 Ricketts and Head Equation with Two Slope Parameters
This is the Ricketts and Head equation with two "slope" parameters (the parameters actually better describe the two different curvatures). It has a different shape than Richards equation and will fit some data sets better. The equation is written in terms of SlopeCon to force Slope1 and Slope2 to be the same sign.If this is not the case then in rare situations, error in the data will result in a fit with slopes of opposite sign.In this case the function attempts to follow the error which results in an irregularly shaped curve. The graph shows the increasing asymmetry with increasing SlopeCon (and therefore increasing Slope2). If SlopeCon=1 then Slope1 = Slope2 and the curve is symmetric and identical to the four parameter logistic curve.
References
Richards, F.J. A flexible growth function for empirical use. J. Exp. Botany 10. pp290-300. Ricketts, J.H. and G. Head. A five-parameter logistic equation for investigating asymmetry of curvature in baroreflex studies. Am. J. Physiol. 277 (Regulatory Integrative Comp. Physiol. 46). R441-R454. 1999.
570 Chapter 12
Survival Curve
This macro computes and graphs a Kaplan-Meier survival curve using the SurvlMod transform. Specify the column containing the survival data, as well as the column indicating censoring of cases. A value of 0 in the censoring column indicates that the case is censored, whereas a value of 1 indicates an uncensored case. Use the Graph Titles section of the Survival Curve dialog box to customize the graph title, the X axis title and the Y axis title. Further customize the plot by selecting between a built-in symbol and a half-line symbol for censored observations. The macro writes the data underlying the plot in seven worksheet columns, beginning in the specified location. The first three result columns contain the time, cumulative probability, and standard error of the cumulative probability. The next two columns hold the coordinates of the censored observations. The final two columns contain the coordinates of the half-lines used to depict the censored observations.
Restrictions
To run the macro, a worksheet must be open and in focus. The worksheet must be sorted by survival times. If identical survival times occur for both censored and uncensored cases, place the censored cases after the uncensored within the tied time value.
Vector Plot
This macro uses the vector transform to plot X,Y, angle and magnitude data as vectors with arrowheads. The arrowheads have a user-specified length and angle. The vector starting point may be selected as the point of the vector tail, midpoint or head. The vector plot consists of three line plots. The data underlying these plots appears in the six columns of the worksheet immediately to the right of the data as three XY pairs.
Restrictions
To run the macro, a worksheet must be open and in focus. The four columns to be plotted must be contiguous and in the following order: X, Y, Angle, and Magnitude (Length). Due to this restriction, only the first data column (X) must be specified. Angle data must be in radians.
Chapter
Using Transforms
13
Transforms are sets of equations that manipulate and calculate data. Math transforms apply math functions to existing data and also generate serial and random data. To perform a transform, you enter variables and standard arithmetic and logic operators into a transform dialog box. Your equations can specify that a transform access data from a worksheet as well as save equation results to a worksheet. You can save transforms as independent .xfm files for later opening or modification. Because transforms are saved as plain text (ASCII) files, you can create and edit them using any word processor that can edit and save text files. This chapter covers: For more information, see User-Defined Transforms below. For more information, see Transform Operators below. For more information, see Performing Quick Transforms below. For more information, see Smoothing 2D and 3D Data below. For more information, see Normalizing Ternary Data below.
Sorting Data
You can use SigmaPlots built-in Sort Selection transform to move selected blocks of data in ascending or descending order according to the order in a key column. Note: Because the sort command sorts data in place, if you want the original data to remain intact, copy the data to a new location and sort the copied data.
571
572 Chapter 13
3. Select the key column by choosing the appropriate column title or column number from the Key Column drop-down list, or by typing the column title or column number in the Key Column box. 4. Select either Ascending or Descending to sort your data in order of increasing or decreasing values. 5. Click OK to sort the data in place and close the Sort Selection dialog box.
General data transformations are math functions and equations which are applied to worksheet data which you can use these to transform data to better fit assumptions of tests, or otherwise modify it before performing a statistical procedure. You can use data transforms to: Generate random numbers. Define dummy variables, lagged variables, and variable interactions. Convert other missing values codes to the "--" double dash symbol for missing values used by SigmaStat.
574 Chapter 13
These commands all appear in the Transforms menu. Choosing these commands prompts you to select the columns to transform, followed by a result column.
Centering Data
The center transform subtracts the mean of a column from all values in that column and places the result in a specified output column. You can often use the center transform on data to eliminate or reduce multicollinearity. For more information on centering data, you can reference any appropriate statistics reference.
To center a variable:
The Pick Columns for Center Transform dialog box appears and prompts you to select an input column. Note:If you select a column in the worksheet before you choose the transform, the selected column is automatically assigned as the input column, and you are prompted for the output column. 2. Pick the worksheet column with the data you want to center as the input column by clicking it in the worksheet or selecting it from the Data for Input drop-down list. The number or title of the selected column appear in the highlighted input row and you are prompted for an output column. 3. Pick the column where you want the centered variables to appear as the output column by clicking it the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appear in the highlighted output row. To change your selections, select the column assignment in the Selected Columnslist, then select the desired column from the worksheet or the drop-down list. You can also double-click a column assignment to clear it. 4. Click Finish to run the transform on the specified input column.
Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
Standardizing Data
Use this transform if you want to standardize variables before performing a statistical procedure. By definition, standardized data has a mean of zero and a standard deviation of one. The standardize transform subtracts the mean of a column from all values in that column, then divides the centered values by the standard deviations.
To standardize a variable:
The Pick Columns for Standardize Transformdialog box appears prompting you to select an Input: column. Note: If you select a column in the worksheet before you choose the transform, the selected column is automatically assigned as the input column in the Selected Columnslist, and you are prompted for the output column. 2. Pick the worksheet column with the data you want to standardize as the input column by clicking it in the worksheet or selecting it from the Data for Input drop-down list. The number or title of the selected column appear in the highlighted input row, and you are prompted for an output column. 3. Pick the column where you want the standardized variables to appear as the output column by clicking it the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row.
576 Chapter 13
4. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 5. Click Finish to run the transform on the specified input column and place the results in the specified output column. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insertto place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
Ranking Data
Use the rank transform to assign integer rank values to data. Ranking data is useful if you want know how the values are ranked, or to perform two way ANOVA on the ranks of data that fails the normality or equal variance tests. The rank transform assigns rank values to all observations in a column from smallest to largest. Equal values are tied in rank, and an averaged rank is assigned to all tied values. This rank is the average of the ranks that would have been assigned to all the tied values if they were not tied.
To rank a variable:
The Data for Rank Transform Pick dialog box appears and prompts you to select an input column. Note: If you select a column in the worksheet before you choose the transform, the selected column is automatically assigned as the input column, and you are prompted for the output column.
2. Pick the column with the data you want to rank as the input column by clicking it in the worksheet or selecting it from the Data for Input drop-down list. The number or title of the selected column appears in the highlighted input row, and you are prompted for an output column. 3. Pick the column where you want the ranked variables to appear as the output column by clicking it the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row. 4. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 5. Click Finish to run the transform on the specified input column and place the results in the specified output column Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
578 Chapter 13
For descriptions of independent variable interactions in multiple linear regression, you can reference any appropriate statistics reference. .
To generate an interaction variable:
The Pick Columns for Interactions Transformdialog box appears and prompts you to select an input column. Note: If you selected columns before you ran the transform, the selected columns are assigned as the input and output columns in the order they were selected in the worksheet. 2. Pick the first variable column with the data you want to factor into the interaction by clicking it in the worksheet or selecting it from the Data for Input drop-down list, then pick the second input column. The number or title of the selected column appears in the input row of the Selected Columns list. 3. Select the column where you want to place the interaction variable as the output column by clicking it in the worksheet or selecting it from the Data for Output dropdown list. The number or title of the selected column appears in the highlighted output row. 4. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 5. Click Finish to run the transform on the specified input columns and place the results in the specified output column. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The
data from the input columns are factored together and placed in the specified output column.
Reference Coding
Reference coding sets the value of all dummy variables to zero when the index variable corresponds to the indexed condition used, and codes all other values of the index variable with a 1.The referenced condition is always assigned a 0. Note: Use reference coding when you want the constant to be the mean of the dependent variable under a selected referenced condition, and the coefficients
580 Chapter 13
computed for the dummy variable(s) to reflect the changes of the constant value from reference condition dependent variable mean.
To create reference coded dummy variables:
1. If necessary, create an index column for your data. These data can consist of any numbers or strings. Each dependent variable value that falls under a different condition is indexed with a different label. For more information, see Indexing Data in Chapter 3. Two factor and repeated measures data require additional index columns. 2. From the menus select:
Transform Statistical Transforms Dummy Variable Reference Coding
The Pick Columns for Reference Transform dialog box appears and prompts you to select input and output columns. Note: If you selected columns before you ran the transform, the selected columns are assigned as the input and output columns in the order they were selected in the worksheet. 3. Pick the column with the indexed data you want to create dummy variables for as the input column by clicking it in the worksheet or selecting it from the Data for Input drop-down list. The number or title of the selected column appear in the highlighted input row and you are prompted for the output column. 4. Select the destination column for the dummy variables as the output column by clicking it in the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row. There should be enough empty columns to the right of the destination column to accommodate all the dummy variable columns; the number of dummy variable columns produced is one less than the number of index values (different groups). 5. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it.
6. Click Finish to run the transform. The Select Reference Index dialog box appears. 7. Select the reference index value from the list to use as the reference condition; no dummy variable is created using this value (this is the condition that determines the constant value; the corresponding dummy variable values for this condition are always zero). All other index values are evaluated for the corresponding dummy variable values. 8. Click OK. The reference coded dummy variables are placed in as many columns as there are index values, less one. Index column values that match the condition used to evaluate the column are assigned a zero; all other values are assigned a 1. One dummy variable column is produced for each index value, except for the index value selected as the reference condition. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
Effects Coding
In effects coding, the dummy variables are coded with -1, 0, and 1. The reference condition is always coded with a -1. The value of other dummy variables is set to zero when the index variable corresponds to the indexed condition used, and set to 1 for all other values of the index variable. Note: Use effects coding when you want the constant term to be computed using the value of the dependent variable under all indexed conditions, and you want the coefficients of the dummy variables to quantify the size of changes from this overall mean.
To create effects coded dummy variables:
1. If necessary, create an index column for your data. This data can consist of any numbers or strings. Each dependent variable value that falls under a different condition
582 Chapter 13
is indexed with a different label. For more information, see Indexing Data in Chapter 3. Two factor and repeated measures data require additional index columns. 2. Choose the Transforms menu Dummy Variables command, and choose Effects Coded. The Pick Columns for Effects Transform dialog box appears and prompts you select an input column. 3. If you selected columns before you ran the transform, the selected columns are assigned as the input and output columns in the order they were selected in the worksheet. 4. Pick the column with the indexed data you want to create dummy variables for as the input column by clicking it in the worksheet or by selecting it from the Data for Input drop-down list. The number or title of the selected column appears in the highlighted input row, and you are prompted for the output column. 5. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 6. Pick the destination column for the dummy variable column(s) as the output column by clicking it in the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row of the Selected Columns list. There should be enough empty columns to the right of the destination column to accommodate all the dummy variable columns; the number of dummy variable columns produced is one less than the number of index values (different groups). 7. Click Finish to run the transform and open the Select Reference Index dialog box. 8. Select the reference index value from the list to use as the reference; no dummy variable is created for this value, and the corresponding dummy variable values for this condition are always -1. All other dummy variable values are set to 1 for the corresponding index variable values. 9. Click OK. Values in the index column that match the index value used to evaluate the column are assigned a zero. Index values that match the reference condition are
assigned -1. All other values are set to 1. One dummy variable column is produced for each index value, except the index selected as the reference condition. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column. If you are creating dummy variables for a two factor or repeated measures problem, create dummy variables for all remaining index columns.
1. Select Multiple Linear Regression from the Standard toolbar drop-down list, then click Run Current Test. 2. Select the dependent variable column, then select the original independent variable and all the dummy variables as the independent variables. 3. Click Finish to run the regression on the selected columns.
584 Chapter 13
4. Compare the results to the original Simple Linear Regression. If the prediction is significantly better, you should consider performing a simple linear regression on the different conditions separately. You can use dummy variables to convert analysis of variance problems into regression problems. For more information on how to do this, you can reference any appropriate statistics reference. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents.The data from the input columns are factored together and placed in the specified output column.
The Pick Columns for Lagged Variable Transform dialog box appears and prompts you to pick an input column. 1. Pick the column with the data you want to lag as the input column by clicking it in the worksheet or selecting it from the Data for Input drop-down list. The number or title of the selected column appears in the highlighted input row of the Selected Columns list, and you are prompted for an output column. 2. Pick the column where you want the lagged variables to appear as the output column by clicking it in the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row of the Selected Columns list.
To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 3. Click Finish to run the transform on the specified input column and place the results in the specified output column. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. 4. Select Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the source column is lagged by one row and placed in the specified column. 5. Repeat this transform if you need to lag the data by additional rows. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
586 Chapter 13
The Filter Data Transform dialog box appears and prompts you to pick a key column. Note: If you select columns before you choose the missing values transform, the selected columns are assigned as the input and output column is the order they were selected in the worksheet. 2. Pick the key column to filter by clicking it in the worksheet or selecting it from the Data for Key drop-down list. This is the column you want to apply the sorting filter to. The number or title of the selected column appears in the highlighted key row, and you are prompted for an output column 3. Pick the column where you want the results of the key column to appear as the output column by clicking it in the worksheet or selecting it from the Data for Output dropdown list. The number or title of the selected column appear in the highlighted output row, and you are prompted for an input column. 4. Pick in the columns that contain the corresponding data to be filtered along with the key column as the input columns, then pick their corresponding output columns by clicking them in the worksheet or selecting them from the drop-down lists. You can pick as many input columns as desired, and you must pick an output column for every input column. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 5. Click Finish to run the Filter transform. The Set Filter dialog box appears. 6. Select Numeric Filter to sort the key column data according to a numeric range. Specify the upper and lower bounds of the values to filter in the Upper Bound and Lower Bound boxes.
7. Select Text Filter to sort the key column data according to a text label in the key column. Enter the string exactly as it appears in the worksheet in the Key Label box and click OK when you have specified the appropriate filter. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. 8. Select Overwrite to replace the existing column contents with the transform results. 9. Select Insert to place transform results above the existing cell contents. Columns are filtered according to the corresponding rows in the key column. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
The Pick Columns for Uniform Random Transform dialog box appears and prompts you for an output column.
588 Chapter 13
Note: Input columns are not selected for the random number transform. 2. Pick the column where you want the random numbers to appear as the output column by clicking it in the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row of the Selected Columns list. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 3. Click Finish to open the Random Number Generation dialog box, and enter the number of random numbers you want to generate in the Quantity box. 4. Enter the lowest and highest numbers in the range of numbers in the Low and High boxes. 5. Enter the seed for the random generator. This is the number used to generate the random numbers. 6. Select Random from the drop-down list to use a random seed number. 7. Click OK when finished. The random numbers are generated according to your specifications and appear in the selected output column. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. 8. Select Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents.
The Pick Columns for Normal Random Transform dialog box appears and prompts you to select an output column. Note: Input numbers are not selected for the random number transform. 2. Pick the column where you want the random numbers to appear as the output column by clicking it in the worksheet or selecting it from the Data for Output drop-down list. The number or title of the selected column appears in the highlighted output row of the Selected Columns list. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 3. Click Finish to open the Normal Random Number Generator dialog box, and enter the number of random numbers you want to generate in the Quantity box. 4. Enter the mean used for the numbers. This is the "middle" or "top" of the bell curve. 5. Enter the standard deviation for the data. The size of this value determines the amount of variation about the mean of the data. A relatively large standard deviation distributes data as a low, flat bell. A relatively small standard deviation creates a tall, skinny bell. 6. Enter the seed for the random number generator. This is the number used to generate the random numbers. Select Random from the drop-down list to use a random seed number. To automatically select a different seed each time the random number generator is used, enter 0/0. 7. Click OK when finished. The random numbers are generated according to your specifications and appear in the selected output. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform
590 Chapter 13
results. Select Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
The Pick Columns for Missing Values Transform dialog box appears and prompts you to select an input column. Note: If you select columns before you choose the missing values transform, the selected columns are assigned as the input and output columns in the order they were selected in the worksheet. 2. Pick the columns with the strings you want convert to missing values as the input column by clicking it in the worksheet or selecting it from the Data for Input dropdown list; then the corresponding output column. You must pick an output column for every input column you select. You can pick as many input columns as desired. The number or title of the selected columns appear in the highlighted input and output rows in the Selected Columns list. 3. To change your selections, select the column assignment in the Selected Columns list, then select the desired column from the worksheet or drop-down list. You can clear a column assignment by double-clicking it. 4. Click Finish to run the transform and open the Missing Value Transform dialog box.
5. Specify the string to replace with missing value symbols. Enter the string exactly as it appears in the worksheet, or select the string from the drop-down list. 6. Click OK when finished. The specified symbols are converted to missing values. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. 7. Select Overwrite to replace the existing column contents with the transform results. Select Insert to place transform results above the existing cell contents. Note: If you specify an output column that contains data, a dialog box appears asking you if you want to erase the column contents, push the contents down, or cancel the transform. Click Overwrite to replace the existing column contents with the transform results. Click Insert to place transform results above the existing cell contents. The data from the input columns are factored together and placed in the specified output column.
592 Chapter 13
Figure 13-3 Selecting Quick Transform from the Transforms dialog box.
2. In the left drop-down list, specify the column or block of cells of the worksheet in which you want the results of the transform to appear.
3. In the right drop-down list, enter a one-line transform using any functions defined in the transform language. You can also use the Functions palette, which appears directly below the Quick Transformsdialog box. Note: You can close the Functions palette by clicking the f(x) button to the right. 4. You cannot edit these equations. To delete an equation, select it from the list and then click the Delete button to the right of the list. Note: You cannot run transforms on date and time columns. To use date and time data, you must first convert the data to numeric data, run the transform, and then convert the column back to date and time data. For more information, see Switching Between Date and Time and Numeric Display in Chapter 3.
1. Click Options on the Quick Transforms dialog box. The Options dialog box appears.
594 Chapter 13
2. Clear Enable automatic updating of transforms. 3. Click OK to save this setting and to close the dialog box. Note: The above instructions explain how to turn off or on automatic updating of transforms specific to a particular worksheet. If you would like to turn this feature on or off globally (the default is set to "on"), you will need to use the main SigmaPlot Options dialog box. For more information, see Setting Program Options in Chapter 1.
The Quick Transform dialog box appears with the Functions palette below, which provides immediate access to frequently used transforms. Using your mouse, you can select cells, blocks of data or columns in the worksheet, and then select different functions on the Functions palette. You can open and close it by clicking the f(x) button. It You can also manually type the equation into the Quick Transform dialog box. For example, if you click Col on the Functions palette, col(?) appears in the Quick Transforms dialog box.
Figure 13-6 Selecting an output column in the Quick Transforms dialog box
Now you can a column number to replace the ?. Or you can first place the cursor in a field in the Quick Transform dialog box, and then select a cell, block of data, or column. This selection then appears in the Quick Transform dialog box.
Figure 13-7 Selecting Column 1 to begin entering transform. The results of the transform will appear in Column 4.
596 Chapter 13
2. To set trigonometric units, click Options. The Options dialog box appears.
Figure 13-8 Click the Options button to set Trigonometric units.
Select the appropriate trigonometric units if calculating trigonometric functions. 3. To use the transform as the title of the column, in the Options dialog box, select Use transform as the title of the output column. For example: Using a Quick Transform of col(3) = col(1)+col(2), results in the column title for column 3 of: col(1)+col(2). Using a Quick Transform of col(4) = col(2)+col(3), results in the column title for column 4 of: col(2)+col(3). 4. Click OK to accept these options and close the dialog box. 5. Click Run to run the transform. Results appear in the cell, block of data, or column specified in the left drop-down list.
2. Select the column with the original X data from the worksheet or the Data Source list. The selected column is assigned as the X Source in the Selected Columns list. 3. Select the Y data source. 4. Select the columns from the worksheet data. 5. Select the X, Y, and Z data destination columns in the worksheet. 6. Click Next. 7. Select the type of scale from the Scale Type drop-down list.
598 Chapter 13
Figure 13-10 Selecting the Scale Type from the Normalize Ternary Data Column Picker Dialog Box
8. Click Finish.
Negative Exponential. Applies a Gaussian weight function to weight the data and a quadratic fit. Use with 2D or 3D data. Bisquare. Applies a bisquare weight function. Use with 2D or 3D data. Inverse Square. Applies a Cauchy weight function. Use with 2D or 3D data. Inverse Distance. Applies the weight function to the (x,y) data. Use with 3D data only. You can find smoother method guidelines in the 2D and 3D Smoothers sections of Samples.jnb. For more information, see About SigmaPlots User and Program Files in Chapter 1.
Smoothing 2D Data
Use the Smooth 2D Data dialog box to remove undesired high-frequency data components, such as data contamination.
To select the data source:
1. Select the worksheet columns by dragging the pointer over your data. 2. From the menus select:
Transforms Smooth 2D Data
600 Chapter 13
3. Click Next.
To select columns for results:
4. Select Predicted: First Empty from the Results list to compute a smoothed value for each data point.
Figure 13-12 Selecting the Results Columns for the Smoothed Data
5. Select Residuals: First Empty to differentiate between the smoothed value and the original Y value. 6. Accept First Next Empty as the standard default column in the Columns drop-down list. 7. Select Plot Results to create a grid of the computed smoothed values on the worksheet. 8. Click Next.
To select columns to graph:
9. Accept First Empty as the default in the Curve Data Column list.
Figure 13-13 Selecting columns to display a grid of smoothed data on the worksheet.
10. Select Create a new graph to create a line plot using the grid of data which appears on the worksheet. 11. To create another plot type and style, clear Create new graph, and create the plot manually. For more information, see Creating 2D Plots in Chapter 6. 12. Click Finish. The Smooth 2D Data dialog box appears.
Figure 13-14 Selecting Smoothers from the Smooth 3D Data drop-down list
13. To define smoothing parameters, select a smoother type from the Smoothers dropdown list.
602 Chapter 13
14. Set the Sampling Proportion to determine a fraction of the total number of data points used to compute each smoothed value. Note: The interpretation of the Sampling Proportion depends on the Bandwidth Method. 15. Set the polynomial degree from the Polynomial Degree list, if applicable. 16. Select Reject Outliers to reduce the effects of outlier points on the smoothed values. 17. To preview and create the graph, click Preview to see a preview of the graph. 18. If the preview is not satisfactory, adjust the Smoother settings and options and click Preview again. Each time you preview, the settings are stored for subsequent review by clicking the right and left arrows. 19. Click OK to accept the preview. The graph appears with a line graph representing the smoothed data points. The original noisy data points also remain. The worksheet now contains the results of all selected computations. Note: You can click the Stop button at the bottom of the Smooth 2D Data dialog box if you want to stop the process.
20. Change the Minimum and Maximum for the X values to new beginning and ending values for the X ranges. For 2D smoothing, the Y values are the smoothed values, and therefore unavailable in the Smoothed Curve Options dialog box. 21. Set the Bandwidth method to either Fixed or Nearest Neighbors. Fixed. Sets the same bandwidth radius the same at every smoothing location. The radius is computed by multiplying the Sampling Proportion value times half of the difference between the set Minimum and Maximum independent variables (X values). Select Fixed if the density of the observed data is relatively constant over the extent of its defined region. Nearest Neighbors. Here the bandwidth radius depends on the smoothing location. The radius is equal to the maximum distance between the smoothing location and its nearest neighbors, as determined by the Sampling Proportion value. Select Nearest Neighbors for data that is clustered in some areas and sparse in others. For example, if there are 100 data points, enter .1 as the Sampling Proportion value to choose ten data points nearest the smoothing location. 22. Click OK to close the dialog box and return to the Smooth 2D Data dialog box.
604 Chapter 13
Smoothing 3D Data
Use the Smoother 3D dialog box to smooth variations in 3D data. You can also resample 3D data to rectangular grid locations to create mesh plots and 3D contour plots from irregularly spaced data. 1. Select the worksheet columns by dragging the pointer over your data. 2. From the menus select:
Transforms Smooth 3D Data
3. Click Next.
Figure 13-17 Selecting the Data Columns to Smooth from the Smoother 3D Dialog Box
4. To select worksheet columns for your results, select Predicted: First Empty from the Results list to compute a smoothed value at each data point. 5. Select Residuals: First Empty to differentiate between the smoothed value and the original Y value. 6. Accept First Empty as the standard default column in the Columns drop-down list. 7. Select Plot Results to create a grid of the computed data on the worksheet. 8. Click Next. 9. To select columns to graph, accept First Empty as the default in the Columns dropdown list.
Figure 13-18 Selecting columns to display the grid of smoothed data
606 Chapter 13
10. Select Create a new graph to create a mesh plot using the grid of data which appears on the worksheet. If you are creating a contour plot, clear Create new graph, and create the contour plot manually. For more information, see Creating Contour Plots in Chapter 7. 11. Select a smoother type from the Smoother drop-down list.
Figure 13-19 Selecting Smoothers from the Smooth 3D Data drop-down list
12. Set the Sampling Proportion, a fraction of a total number of data points used to compute each smoothed value. Note: The Sampling Proportion depends on the Bandwidth Method. 13. Set the Polynomial Degree from the Polynomial Degree list, if applicable. 14. Select Reject Outliers to reduce the effects of outlier points on the smoothed values. 15. To set the smoothed surface options, click Options. The Smoothed Surface Options dialog box appears. For more information, see Setting Smoothed Surface Options below. 16. To preview and then create the graph, click Preview to see a preview of the graph. If the preview is not satisfactory, adjust the Smoother settings and options, and click Preview again. Each time you preview, the settings are stored for subsequent review by clicking the right and left arrows.
17. Click OK to accept the preview. The graph appears, and the worksheet now contains the results of all selected computations. Note: You can click the red Stop button at the bottom of the Smooth 3D Data dialog box to stop the process.
18. Change the Minimum and Maximum for the X and Y values to new beginning and ending values for the X and Y ranges. 19. Set the bandwidth method to either Fixed or Nearest Neighbors. Fixed. The bandwidth radius is the same at every smoothing location. The radius is computed by multiplying the Sampling Proportion value times half of the difference between the set Minimum and Maximum independent variables (X and Y values). Select Fixed if the density of the observed data is relatively constant over the extent of its defined region.
608 Chapter 13
Nearest Neighbors. Here the bandwidth radius depends on the smoothing location. The radius is equal to the maximum distance between the smoothing location and its nearest neighbors, as determined by the Sampling Proportion value. Select Nearest Neighbors for data that is clustered in some areas and sparse in others. 20. Click OK to close the dialog box and return to the Smooth 3D Data dialog box.
User-Defined Transforms
Modify and manipulate worksheet data by entering SigmaPlots extensive mathematical transformation language into the User-Defined Transform dialog box. Use transforms to create new data by performing functions on existing data, or generate calculated or random data, which can then be placed in worksheet columns. For more information, see Transform Operators below. The first step to transform worksheet data is to enter the desired equations in the edit box of the User-Defined Transform dialog box. If no previously entered transform equations exist, the edit box is empty; otherwise, the last transform entered appears. Select the edit box to begin entering transform instructions. As you enter text into the transform edit box, the box scrolls down to accommodate additional lines. You can enter up to 100 lines of equations, either on separate lines or on the same line.
3. Type transform instructions into the Edit Transform field. You can enter up to 32,000 characters. 4. Click Run. You can save the contents of the transform window to a file. Since this is a text file, you can view or print these files using any word processor. You can open previously saved transforms in the transform window for execution or modification. All transform files have the extension of .xfm in the Transforms folder. To view these files, click the Open button in the User-Defined Transforms dialog box and open a transform file. A library of transform results is named Xfms.jnb in the Transform folder. These transform examples also include a sample SigmaPlot graph file displaying the results of the transform. For more information, see About SigmaPlots User and Program Files in Chapter 1.
610 Chapter 13
Transform Syntax
Enter transforms as equations with the results placed to the left of the equal sign (=) and the calculation placed to the right of the equal sign. Results can be defined as either variables (which can be used in other equations), or as the worksheet column or cells where results are to be placed.
Entering Transforms
To type an equation in the transform edit box, click in the edit box and begin typing. When you complete a line, press Enter to move the cursor to the first position on the next line. You can leave spaces between equation elements: x = a+b is the same as x = a + b. However, you may find it necessary to conserve space by omitting spaces. Blank lines are ignored so that you can use them to separate or group equations for easier reading.
Figure 13-22 Typing Equations into the Edit Window
If the equation requires more than one line, you may want to begin the second and any subsequent lines indented a couple of spaces (press the space bar before typing the line). Although this is not necessary, indenting helps distinguish a continuing equation from a new one. Note: You can resize the transform dialog box to enlarge the edit box. You can press Ctrl+X, Ctrl+C, and Ctrl+V to cut, copy, and paste text in the edit window.
Transforms are limited to a maximum of 100 lines. Note that you can enter more than one transform statement on a line; however, this is only recommended if space is a premium. Note: Use only parentheses to enclose expressions. Curly brackets and square brackets are reserved for other uses.
Commenting on Equations
To enter a comment, type an apostrophe () or a semicolon (;), then type the comment to the right of the apostrophe or semicolon. If the comment requires more than one line, repeat the apostrophe or semicolon on each line before continuing the comment.
Sequence of Expression
SigmaPlot and SigmaStat generally solve equations regardless of their sequence in the transform edit box. However, the col function (which returns the values in a worksheet column) depends on the sequence of the equations, as shown in the following example.
Example
The sequence of the equations: col(1)=col(4)^alpha col(2)=col(1)*theta must occur as shown. The second equation depends on the data produced by the first. Reversing the order produces different results. To avoid this sequence problem, assign variables to the results of the computation, then equate the variables to columns: x=col(4) y=x^alpha z=y*theta col(1)=y col(2)=z The sequence of the equations is now unimportant.
612 Chapter 13
Transform Components
Transform equations consist of variables and functions. Operators are used to define variables or apply functions to scalars and ranges. A scalar is a single worksheet cell, number, missing value, or text string. A range is a worksheet column or group of scalars.
Figure 13-23 Examples of the Transform Equation Elements Typed into the Transform Window
Variables
You can define variables for use in other equations within a transform. Variable definition uses the following form: variable = expression Variable names must begin with a letter. After that, they can include any letter or number, or the underscore character ( _ ). Variable names are case sensitivean "A" is not the equivalent of an "a." Once a variable has been defined by means of an expression, that variable cannot be redefined within the same transform.
Functions
A function is similar to a variable, except that it refers to a general expression, not a specific one, and thus requires arguments. The syntax for a function declaration is function(argument 1,argument 2,...) = expression
where function is the name of the function, and one or more argument names are enclosed in parentheses. Function and argument names must follow the same rules as variable names. User-Defined Functions. Frequently used functions can be copied to the Clipboard and pasted into the transform window.
Constructs
Transform constructs are special structures that allow more complex procedures than functions. Constructs begin with an opening condition statement, followed by one or more transform equations, and end with a closing statement. The available constructs are for loops and if...then...else statements.
Operators
A complete set of arithmetic, relational, and logic operators are provided. Arithmetic operators perform simple math between numbers. Relational operators define limits and conditions between numbers, variables, and equations. Logic operators set simple conditions for if statements. For more information, see Transform Operators below.
Numbers
You can enter numbers as integers, in floating point style, or in scientific notation. All numbers are stored with 15 figures of significance. Use a minus sign in front of the number to signify a negative value. Missing values, represented in the worksheet as a pair of dashes, are considered non-numeric. All arithmetic operations which include a missing value result in another missing value. To generate a missing value, divide zero by zero.
Example
614 Chapter 13
size({1,2,3,missing}) returns a value of 4.0. (The size function returns the number of elements in a range, including labels and missing values.) The transform language does not recognize two successive dashes; for example, the string {1,2,3,--} is not recognized as a valid range. Dashes are used to represent missing values in the worksheet only. Strings, such as text labels placed in worksheet cells, are also non-numeric information. To define a text string in a transform, enclose it with double quotation marks. As with missing values, strings may not be operated upon, but are propagated through an operation. The exception is for relational operators, which make a lexical comparison of the strings, and return true or false results accordingly.
The operation:
{col(1), col(2)} concatenates columns 1 and 2 into a single range. Note that elements constituting a range need not be of the same type, i.e., numbers, labels and missing values.
Example 3
Array References
Individual scalars can be accessed within a range by means of the square bracket ([ ]) constructor notation. If the bracket notation encloses a range, each entry in the enclosed range is used to access a scalar, resulting in a new range with the elements rearranged.
Example
For the range: x = {1.4,3.7,3.3,4.8} the notation: x[3] returns 3.3, the third element in the range. The notation: x[{4,1,2}] produces the range {4.8,1.4,3.7}. The constructor notation is not restricted to variables: any expression that produces a range can use this notation.
Example
616 Chapter 13
produces the same result as col(3,2,2), or cell(3,2). The notation: {2,4,6,8}[3] produces 6. If the value enclosed in the square brackets is also a range, a range consisting of the specified values is produced.
Example
The operation: col(1)[{1,3,5}] produces the first, third, and fifth elements of column 1.
Figure 13-24 Range and Array Reference Operations Typed into the User Defined Transform Window
Transform Operators
Transforms use operators to define variables and apply functions. A complete set of arithmetic, relational, and logical operators are provided.
Order of Operation
The order of precedence is consistent with P.E.M.A. (Parentheses, Exponentiation, Multiplication, and Addition) and proceeds as follows, except that parentheses override any other rule: Exponentiation, associating from right to left. Unary minus. Multiplication and division, associating from left to right. Addition and subtraction, associating from left to right. Relational operators. Logical negation. Logical and, associating from left to right. Logical or, associating from left to right. This list permits complicated expressions to be written without requiring too many parentheses.
Figure 13-25 Examples of Transform Operators
Example
The statement:
618 Chapter 13
a<10 and b<5 groups to (a<10) and (b<5), not to (a<(10 and b))<5. Note: Only parentheses can group terms for processing. Curly and square brackets are reserved for other uses.
Operations on Ranges
The standard arithmetic operatorsaddition, subtraction, multiplication, division, and exponentiationfollow basic rules when used with scalars. For operations involving two ranges corresponding entries are added, subtracted, etc., resulting in a range representing the sums, differences, etc., of the two ranges. If one range is shorter than the other, the operation continues to the length of the longer range, and missing value symbols are used where the shorter range ends. For operations involving a range and a scalar, the scalar is used against each entry in the range. Example: The operation: col(4)*2 produces a range of values, with each entry twice the value of the corresponding value in column 4.
Arithmetic Operators
Arithmetic operators perform arithmetic between a scalar or range and return the result. + * / Add Subtract (also signifies unary minus) Multiply Divide
^ or ** Exponentiate Multiplication must be explicitly noted with the asterisk. Adjacent parenthetical terms such as (a+b) (c-4) are not automatically multiplied.
Relational Operators
Relational operators specify the relation between variables and scalars, ranges or equations, or between user-defined functions and equations, establishing definitions, limits and/or conditions. = or .EQ.Equal to > or .GT.Greater than >= or .GE.Greater than or equal to < or .LT.Less than <= or .LE.Less than or equal to <>, !=, #, or .NE.Not equal to The alphabetic characters can be entered in upper or lower case.
620 Chapter 13
Logical Operators
Logical operators are used to set the conditions for if function statements. and, & Intersection or, | not, ~ Union Negation
Transform Examples
Many mathematical transform examples, along with appropriate graphs and worksheets are included with SigmaPlot. This chapter is describes the data transform examples and the graphing transform examples provided. Each description contains the text of the transform and, where applicable, a graph displaying the possible results of the transform. You can find these sample transforms in the Transforms folder. For more information, see About SigmaPlots User and Program Files in Chapter 1.
1. Make sure your original Y data is in column 2. Perform the desired regression using the Regression Wizard, and save your Predicted values (fitted Y data) in column 3, and Parameters (the regression coefficients) in column 4. 2. Press F10 to open the User-Defined Transform dialog box, then click the Open button and open the ANOVA.XFM transform file in the XFMS directory. The ANOVA transform appears in the edit window. 3. Click Run. The ANOVA results are placed in columns 5 through 9, or beginning at the column specified with the anova variable.
622 Chapter 13
1. Place your X data in column 1 and your Y data in column 2. If your data has been placed in other columns, you can specify these columns after you open the AREA.XFM file. You can use an existing or new worksheet. 2. Press F10 to open the User-Defined Transform dialog box, then click the Open button and open the AREA.XFM transform file in the XFMS directory. The Area transform appears in the edit window. 3. Click Run. The area is placed in column 3 or in the column specified with the res variable.
Bivariate Statistics
This transform takes two data columns of equal length and computes their means, standard deviations, covariance, and correlation coefficient. The columns must be of equal length. The Bivariate transform uses examples of these transform functions mean stddev total To use the Bivariate transform: 1. Place your X data in column 1 and your Y data in column 2. If your data has been placed in other columns, you can specify these columns after you open the
BIVARIAT.XFM transform file. You can enter data into an existing worksheet or a new worksheet. 2. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the BIVARIAT.XFM transform file in the XFMS directory. The Bivariate Statistics transform appears in the edit window. 3. Click Run. The results are placed in columns 3 and 4, or beginning in the column specified with the res variable.
1. Begin a new worksheet by choosing the File menu New command, then choosing Worksheet; this transform requires a clean worksheet to work correctly. 2. Open the User-Defined Transforms dialog box by selecting the Transforms menu User Defined command, then clicking the Open button, and opening the DIFFEQN.XFM transform file in the XFMS directory. The Differential Equation Solving transform appears in the edit window. 3. Scroll to the Number of Equations section and enter a value for the neqn variable. This is the number of equations you want to solve, up to four. 4. Scroll down to the Differential Equations section, and set the fp1 through fp4 functions to the desired functions. Set any unused equations = 0. If only one first order differential equation is used, then only the fp1 transform equation is used and fp2, fp3, and fp4 are set to 0. For example, if you only wanted to solve the differential equation:
624 Chapter 13
fp1(x,y1,y2,y3,y4) = -a*y1 fp2(x,y1,y2,y3,y4) = 0 fp3(x,y1,y2,y3,y4) = 0 fp4(x,y1,y2,y3,y4) = 0 5. Scroll down to the Initial Values heading and set the nstep variable to the number of integration (X variable) steps you want to use. The more steps you set, the longer the transform takes. 6. Set the initial X value x0, final X value x1, and the Y1 through Y4 values (placed in cells (2,1) through (5,1)). If you are not using a y1 value, set that value to zero (0). For example, for the single equation example above, you could enter: x0 = 0 ;initial x x1 = 1 ;final x cell(2,1) = 1 ;y1 initial value cell(3,1) = 0 ;y2 initial value cell(4,1) = 0 ;y3 initial value cell(5,1) = 0 ;y4 initial value
7. Click Run. The results output is placed in columns 1 through neqn+1. 8. To graph your results, create a Line Plot graphing column 1 as your X data and columns 2 through 5 as your Y data.
626 Chapter 13
1. To use the provided sample data and graph, open the F-test worksheet and graph in the XFMS.JNB notebook. The worksheet contains raw data in columns 1 and 2, and curve fit results for the two competitive binding models in columns 3-5 and 6-8. The graph plots the raw data and the two curve fits. 2. To use your own data, enter the XY data to be curve fit in columns 1 and 2, respectively. Select the first curve fit equation and use it to fit the data, place the parameters, fit results and residuals in the first empty columns (3-5). Run the second curve fit and place the results in columns 6-8 (the default). If desired, create graphs of these results using the wizard. 3. Press F10 to open the User-Defined Transform dialog box, then open the F_TEST.XFM transform file. Specify n1 and n2, the number of parameters in the lower and higher order functions. In the example provided, these are 3 and 5, respectively. If necessary, specify cs1 and cs2, the column locations for the residuals of each curve fit, and cres, the first column for the two column output. 4. Click Run. The F-test value and corresponding P value are placed into the worksheet. If P < 0.05, you can predict that the higher order equation provides a statistically better fit.
Figure 13-29 Comparing Two Curve Fits
1. Place the X data in column 1 of the worksheet and the Y data in column 2. If your data is in other columns, you can specify these columns after you open the STDV_REG.XFM transform file. You can enter data into an existing worksheet or a new worksheet. 2. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the STDV_REG.XFM transform file in the XFMS directory. If necessary, change the x_col, y_col, and res variables to the correct column numbers.
628 Chapter 13
3. Click Run. The results are placed in columns 3 and 4, or in the columns specified by the res variable.
3. Press F10 to open the User-Defined Transform dialog box, then click the Open button and open the CONTCHRT.XFM transform file in the XFMS directory. The Control Chart transform appears in the edit window. 4. Click Run. The results are placed in columns 4 through 5 of the worksheet. 5. If you opened the Control Chart graph, view the graph page. The graph plots the fraction of defectives using a Line and Scatter plot with a Simple Straight Line style graphing column 3 as Y data versus the row numbers. The control lines are plotted as a Simple Horizontal Step Plot using columns 4 and 5 versus their row numbers. The mean line for the fractional defectives is drawn with a reference line. 6. To create your own graph, create a Line and Scatter Plot, with a Simple Line style, then plot column 3 as Y data against the row numbers. Add an additional Line Plot using the Multiple Horizontal Step Plot style, plotting columns 4 and 5 versus their two numbers, then add a reference line to plot the mean line for the fractional device.
Figure 13-30 Control Chart Graph
630 Chapter 13
ond Derivatives
This example takes data with irregularly spaced X values and generates a cubic spline interpolant. The CBESPLN1.XFM transform takes X data which may be irregularly spaced and generates the coefficients for a cubic spline interpolant. The CBESPLN2.XFM transform takes the coefficients and generates the spline interpolant and its two derivatives. The values for the interpolant start at a specified minimum X which may be less than, equal to, or greater than the X value of the original first data point. The interpolant has equally spaced X values that end at a specified maximum which may be less than, equal to, or greater than the largest X value of the original data. Note that this is not the same algorithm that SigmaPlot uses; this algorithm does not handle multiple valued functions, whereas SigmaPlot does. To use the transform to generate and graph a cubic spline interpolant, you can either use the provided sample data and graph, or begin a new notebook, enter your own data and create your own graph using the data. 1. To use the provided sample data and graph, open the Cubic Spline worksheet and graph by double-clicking the graph page icon in the Cubic Spline section of the Transform Examples notebook. The worksheet appears with data in columns 1 and 2 and the graph page appears with two graphs. The first graph plots the original XY data as a scatter plot. The second graph appears empty. 2. To use your own data, enter the irregularly spaced XY data into the worksheet. The X values must be sorted in strictly increasing values. The default X and Y data columns used by the transform are columns 1 and 2, respectively. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the CBESPLN1.XFM transform file in the XFMS directory. The first Cubic Spline transform appears in the edit window. 4. Move to the Input Variables heading. Set the X data column variable cx, the Y data column cy, the beginning interpolated X value xbegin, the ending interpolated X value xend, and the X increments for the interpolated points xstep. A larger X step results in a smoother curve but takes longer to compute. Enter the end condition setting iend for the interpolation. 5. Enter the end condition setting iend for the interpolation.
You can use first, second, or third order conditions. If you have only a few data points, you should try different orders to see which one you like the most. See the example for the effect of too low an order on the first and second derivatives. 1 end spline segments approach straight lines asymptotically 2 end spline segments approach parabolas asymptotically 3 end spline segments approach cubics asymptotically 6. Move to the RESULTS heading and enter the first column number for the results cr. This column for the beginning of the results block is specified in both transforms. 7. Click Run to run the transform. When it finishes, press F10 then open the CBESPLN2.XFM transform file in the XFMS directory. Make sure that the cr variable is identical to the previous value, then click Run. 8. If you opened the Cubic Spline graph, view the page. The first graph plots the original XY data as a scatter plot and the interpolated data as a second line plot by picking the cr column as the X column and cr+1 as the Y column. The second graph plots the derivatives as line plots using the cr column versus the cr+2 column and the cr column versus the cr+3 column. 9. To create your own graphs using SigmaPlot, create a Scatter Plot using a Simple Scatter style which plots the original data in columns 1 and 2 as XY pairs. Add an additional Line Plot using a Simple Spline Curve, then plot the cr column as the X column against the cr+1 column as the Y column.
632 Chapter 13
Figure 13-31
r1 i1 r2 i2 .... .... rn in where r values are real elements, and i values are imaginary elements. In transform language syntax, the two columns {{r1, r2, ... rn},{i1, i2, ... in}} are written as: block({r1, r2, ... rn},{i1, i2, ... in}) This function works on data sizes of size 2n numbers. If your data set is not 2n in length, the fft function pads 0 at the beginning and end of the data range to make the length 2n. For more information, see Smoothing with a Low Pass Filter. The fft function returns a range of complex numbers. The Fast Fourier Transform is usually graphed with respect to frequency. To produce a frequency scale, use the relationship: f=fs*(data(0,n/2)-1)/n where fs is the sampling frequency. The example transform POWSPEC.XFM. includes the automatic generation of a frequency scale. The Fast Fourier Transform operates on data which is assumed to be periodic over the interval being analyzed. If the data is not periodic, then unwanted high frequency components are introduced. To prevent these high frequency components from occurring, windows can be applied to the data before using the fft transform. The Hanning window is a cosine function that drops to zero at each end of the data. The example transform POWSPEC.XFM includes the option to implement the Hanning window. For more information, see Computing Power Spectral Density below.
634 Chapter 13
1. First assign the data you want to filter to column 1 of the worksheet. You can generate the data using a transform, or use your own measurements. 2. Press F10 to open the User-Defined Transforms dialog box, then click the New button to start a new transform. 3. Type the following transform in the edit window: x=col(1) real data tx=fft(x) compute the fft block(2)=tx place real fft data back in col(2) place imaginary fft data in col(3) 4. Click Run. The results are placed starting one column over from the original data. Computing Power Spectral Density The example transform POWSPEC.XFM uses the Fast Fourier Transform function, then computes the power spectral density, a frequency axis, and makes optional use of a Hanning window. To calculate and graph the power spectral density of a set of data, you can either use the provided sample data and graph, or begin a new notebook, enter your own data and create your own graph using the data. 5. To use the sample worksheet and graph, open the Power Spectral Density worksheet and graph by double-clicking the graph page icon in the Power Spectral Density section of the Transform Examples notebook. Data appears in column 1 of the worksheet, and two graphs appear on the graph page. The top graph shows data generated by the sum of two sine waves plus Gaussian random noise. The data is represented by: f(t)=sin(2*pi*f1*t)+0.3*sin(2*pi*f2*t)+g(t) where f1=10 cycles/sec (cps), f2=100cps, and the Gaussian random noise has mean 0 and standard deviation of 0.2. The lower graph is empty. 6. To use your own data, place your data in column 1. If your data is in a different column, specify the new column after you open the POWSPEC.XFM transform file.
7. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open POWSPEC.XFM transform file in the XFMS directory. The Power Spectral Density transform appears in the edit window. Note: To use this transform, the Trigonometric Units must be set to Radians. 8. Click Run. Since the frequency sampling value (fs) is nonzero, a frequency axis is generated in column 2 and the power spectral density data in column 3. 9. If you opened the Power Spectral Density graph, view the graph page. Two graphs appear on the page. The top graph plots the data generated by the sum of two sine waves plus Gaussian random noise using a Line Plot with Simple Straight Line style graphing column 1 versus row numbers. The lower graph plots the power spectral density using a Line Plot with a Simple Straight Line style, graphing column 2 as the X data (frequency), and column 3 as the Y data. 10. To plot your own data using SigmaPlot, choose the Graph menu Create Graph command, or select the Graph Wizard from the toolbar. Create a Line Plot with a Simple Straight Line style plotting your original data versus row numbers by choosing Single Y data format. If you set the frequency sampling value (fs) to nonzero, create a Line Plot with a Simple Straight Line style, graphing columns 2 and 3 using XY Pair data format. Otherwise, create a Line Plot with a Simple Straight Line style plotting column 3 (power spectral density) versus row numbers by choosing Single Y data format. The power spectral density plot of the signal f(t) shows two major peaks at the two frequencies of the sine waves (10cps and 100cps), and a more or less constant noise level in between. For more information, see Creating and Modifying Graphs in Chapter 4.
636 Chapter 13
Kernel Smoothing
The example transform SMOOTH.XFM smooths data by convolving the Fast Fourier Transform of a triangular smoothing kernel together with the fft of the data. Smoothing data using this transform is computationally very fast; the number of operations is greatly reduced over traditional methods, and the results are comparable. To increase the smoothing, increase the width of the triangular smoothing kernel. To calculate and graph the smoothed data, you can either use the provided sample data and graph, or begin a new notebook, enter your own data, and create your own graph using the data. 1. To use the sample worksheet and graph, open the Kernel Smoothing worksheet and graph by double-clicking the graph page icon in the Kernel Smoothing section of the Transform Examples notebook. Data appears in columns 1 through 4, 6, and 7 of the worksheet, and two graphs appear on the graph page. The first graph has two plots, the signal, and the signal with noise distortion. Column 1 contains the X data, column 2
contains the Y data for the signal, and column 3 contains the Y data for the signal and the noise distortion. The lower graph is empty. 2. To use your own data, place your data in columns 1 through 2. If your data is in other columns, specify the new columns after you open the SMOOTH.XFM transform file. If necessary, specify a new column for the results. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open SMOOTH.XFM transform file in the XFMS directory. The Kernel Smoothing transform appears in the edit window. Note: To use this transform, make sure the Insert mode is turned off. 4. Click Run. The results are placed in column 5 unless you specified a different column in the transform. 5. If you opened the Kernel Smoothing graph, view the graph page. Two graphs appear on the page. The first graph has two plots, the signal, and the signal with noise distortion. The Line Plot with a Multiple Straight Line style graphs column 1 as the X data, column 2 as the Y data for the signal, and column 3 as the Y data for the signal and the noise distortion. The lower Line Plot with a Simple Straight Line style plots column 1 as the X data, and column 5 as the Y data using XY Pairs data format. 6. To plot your own data using SigmaPlot, choose the Graph menu Create Graph command, or select the Graph Wizard from the toolbar. Create a Line Plot with a Multiple Straight Line style using X Many Y data format, plotting column 1 as the X data, column 2 as the Y data for the signal, and column 3 as the Y data for the signal and the noise distortion. Create a second Line Plot graph with a Simple Straight Line style using the data in columns 1 and 5, graphing column 1 as the X data and column 5 as the Y data using XY Pairs data format. For more information, see Creating and Modifying Graphs in Chapter 4.
638 Chapter 13
The LOWPASS.XFM transform expresses f as a percentage for ease of use. As the value of f increases, more high frequency channels are removed. Note that this is a digital transform which cuts data at a discrete boundary. In addition, this transform does not alter the phase of the data, which makes it more accurate than analog filtering. A high pass or band pass filter can be constructed in the same manner. To calculate and graph the smoothing of a set of data using a low pass filter, you can either use the provided sample data and graph, or begin a new notebook, enter your own data, and create your own graph using the data. 1. To use the sample worksheet and graph, open the Low Pass Smoothing worksheet and graph by double-clicking the graph page icon in the Low Pass Smoothing section of the Transform Examples notebook. Data appears in columns 1 through 4 of the worksheet, and two graphs showing plots appear on the graph page. Column 1 contains the X data, column 2 contains the Y data for the signal and the noise distortion, column 3 contains the X data, and column 4 contains the Y data for the original signal. The top graph plots the signal plus the noise distortion; the bottom graph plots the signal. 2. To use your own data, place your data in columns 1 through 2. If your data is in other columns, specify the new columns after you open the LOWPASS.XFM transform file. If necessary, specify a new column for the results. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open LOWPASS.XFM transform file in the XFMS directory. The Low Pass Filter transform appears in the edit window. Note: To use this transform, make sure Insert mode is turned off. 4. Click Run. The results are placed starting in column 5, unless you specified a different column in the transform. 5. If you opened the Low Pass Smoothing graph, view the graph page. Two graphs appear. The top graph plots the signal plus the noise distortion, using a Line Plot with a Simple Straight Line style and XY Pairs data format graphing column 1 as the X data, column 2 as the Y data for the signal and the noise distortion. The bottom graph displays two plots. A Scatter Plot with a Simple Scatter Style and XY Pairs data format, plots column 3 as the X data, and column 4 as the Y data for the original signal. A second Line Plot with a Simple Straight Line style using data in columns 1 and 5, plots column 1 as the X data and column 5 as the Y data using XY Pairs data format.
640 Chapter 13
6. To plot your own data using SigmaPlot, choose the Graph menu Create Graph command, or select the Graph Wizard from the toolbar. Create two graphs. Graph the signal plus the noise distortion, using a Line Plot with a Simple Straight Line style and XY Pairs data format graphing column 1 as the X data, column 2 as the Y data for the signal and the noise distortion. Create a second graph with two plots. Plot the original signal using a Scatter Plot with a Simple Scatter Style and XY Pairs data format, plotting column 3 as the X data, and column 4 as the Y data for the original signal. Add a second Line Plot with a Simple Straight Line style using data in columns 1 and 5, plotting column 1 as the X data and column 5 as the Y data using XY Pairs data format.
Figure 13-34 Low Pass Filter Smoothing Graph
tx=fft(x) compute fft of data md=real(tx)^2+img(tx)^2 compute sd kc=if(md>P,1,0) remove frequencies with psd<P sd=mulcpx(complex(kc),tx) remove frequency components from x td=real( invfft(sd) ) convert back to time domain col(2)=td place results in worksheet To calculate and graph the smoothing of a set of data using a gain filter, you can either use the provided sample data and graph, or begin a new notebook, enter your own data, and create your own graph using the data. 1. To use the sample worksheet and graph, open the Gain Filter Smoothing worksheet and graph by double-clicking the graph page icon in the Gain Filter Smoothing section of the Transform Examples notebook. Data appears in columns 1 through 3 of the worksheet, and two graphs showing plots, and one blank graph appear on the graph page. Column 1 contains the Y data for the signal plus noise, column 2 contains the X data and column 3 contains the Y data for the power spectral density graph. The top graph plots the signal plus the noise distortion; the middle graph plots the power spectral density. 2. To use your own data, place your data in column 1. If your data is in a different column, specify the new column after you open the GAINFILT.XFM transform file. If necessary, specify a new column for the results. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open GAINFILT.XFM transform file in the XFMS directory. The Gain Filter transform appears in the edit window. Note: To use this transform, make sure Insert mode is turned off. For more information, see Insertion and Overwrite Modes in Chapter 3. 4. Click Run. The results are placed in column 5 unless you specified a different column in the transform. 5. If you opened the Gain Filter Smoothing graph, view the graph page. Three graphs appear. The top graph plots the signal plus the noise distortion using a Line Plot with
642 Chapter 13
a Simple Straight line style and Single Y data format, plotting column 1 as the Y data for the signal plus noise. The middle graph plots the power spectral density using a Line Plot with a Simple Straight Line style and XY Pairs data format, plotting column 2 as the X data and column 3 as the Y data for the power spectral density graph. The lower graph is a plot of the gain filtered signal, using a Line Plot with a Simple Straight Line style, and single Y data format from column 5. 6. To plot your own data using SigmaPlot, choose the Graph menu Create Graph command, or select the Graph Wizard from the toolbar. Create two graphs. Plot the signal plus the noise distortion using a Line Plot with a Simple Straight line style and Single Y data format, plotting column 1 as the Y data for the signal plus noise. Plot the gain filtered signal using a Line Plot with a Simple Straight Line style, and single Y data format from column 5.
Frequency Plot
This transform example creates a frequency plot showing the frequency of the occurrence of data in the Y direction. Data is grouped in specified intervals, then horizontally plotted for a specific Y value. Parameters can be set to display symbols that are displaced a specific distance from each other or that touch or overlap. You can also plot the mean value of each data interval. This transform example shows overlapping symbols which give the impression of data mass.
644 Chapter 13
To calculate and graph the frequency of the occurrence of a set of data, you can either use the provided sample data and graph, or begin a new notebook, enter your own data and create your own graph using the data. 1. To use the sample worksheet and graph, open the Frequency Plot worksheet and graph by double-clicking the graph page icon in the Frequency Plot section of the Transform Examples notebook. Data appears in columns 1 through 3 of the worksheet, and an empty graph appears on the graph page. 2. To use your own data, place your data in columns 1 through 3. You can put data in as many or as few columns as desired, but if you use the sample transform you must change the X locations of the Y values in the second line under the Input heading in the transform file to reflect the number of data columns you are using. If your data is in other columns or more than three columns, specify the new columns after you open the FREQPLOT.XFM transform file. Enter the tick labels for the X axis in a separate column, and specify tick labels from a column using the Tick Labels Type drop-down list in the Tick Labels panel in Graph Properties Axis tab. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the FREQPLOT.XFM transform file in the XFMS directory. The Frequency Plot transform appears in the edit window. 4. Click Run. The results are placed starting one column over from the original data. 5. If you opened the sample Frequency Plot graph, view the graph page. A Scatter Plot appears plotting columns 5 and 6, 7 and 8, and 9 and 10 as three separate XY Pair plots. The lines passing through each data interval is a fourth Line Plot with a Simple Straight Line style plotting columns 11 and 12 as an XY pair, representing the mean value of each data interval. The X axis tick marks are generated by the transform. The axis labels are taken from column 13.
6. To create your own graph using SigmaPlot, make a graph with three Scatter Plots with Simple Scatter styles. Plot each consecutive result column pair as XY pair scatter plots. If the mean line option is active in the transform, plot the last consecutive result column pair as a XY pair Line Plot with Simple Straight Line style. Use labels typed into a worksheet column as the X axis tick labels.
646 Chapter 13
Examples notebook. Data appears in column 1 of the worksheet and two empty graphs appear on the graph page. 2. To use your own data, place the X data in column 1. If your data has been placed in another column, you can specify the column after you open the GAUSDIST.XFM transform file. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the GAUSDIST.XFM transform file in the XFMS directory. The Gaussian Cumulative transform appears in the edit window. 4. Click Run. The results are placed in column 2, or in the column specified by the res variable. 5. If you opened the sample Gaussian graph, view the graph page. A Line Plot appears with a spline curve in the first graph with column 1 as the X data versus column 2 as the distribution (Y) data. 6. To create your own graph using SigmaPlot, make a Line Plot graph with a Simple Spline Curve. The spline curve plots column 1 as the X data versus column 2 as the distribution (Y) data.
1. Open the Histogram Gaussian worksheet and graph by double-clicking the graph page icon in the Histogram Gaussian section of the Transform Examples notebook.
648 Chapter 13
The Histogram worksheet with data in column 1 and an empty graph page appears. The data in the Histogram Gaussian worksheet was generated using the transform: col(1) = gaussian(100,0,325,2)
To use your own data:
2. Place the sample in column 1 of the worksheet. If your data has been placed in another column, you can specify this column after you open the HISTGAUS.XFM transform file. You can enter the data into an existing or new worksheet. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the HISTGAUS.XFM transform file in the XFMS directory. The Histogram with Gaussian Distribution transform appears in the edit window. 4. Click Run. The results are placed in columns 2 through 5 of the worksheet, or in the columns specified by the res variable. 5. If you opened the Histogram Gaussian graph, view the graph page. A histogram appears using column 2 as X data versus column 3 as the Y data. The curve plots the Gaussian distribution using column 4 as X data versus column 5 as the Y data. 6. To create your own graph using SigmaPlot, create a simple vertical bar chart and set the bar widths as wide as possible. Add the Gaussian curve to the graph by creating another plot using the data in column 4 as the X data and the data in column 5 as the Y data.
650 Chapter 13
The graph page appears with a scatter graph plotting the original data in columns 1 and 2. 2. To use your own data, place the X data in column 1 and the Y data in column 2. If your data has been placed in other columns, you can specify these columns after you open the LINREGR.XFM transform file. You can enter data into an existing or a new worksheet. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the LINREGR.XFM transform in the XFMS directory. The Linear Regression transform appears in the edit window. If necessary, change the x_col, y_col, and res variables to the correct column numbers (this is not necessary for the example Linear Regression worksheet data). 4. Change the Z variable to reflect the desired confidence level (this is not necessary for the example Linear Regression worksheet data). 5. Click Run. The results are placed in columns 3 through 8, or in the columns specified by the res variable. 6. If you opened the Linear Regression graph, view the graph page. The original data in columns 1 and 2 is plotted as a scatter plot. The regression is plotted as a solid line plot using column 3 as the X data versus column 4 as the Y data, the confidence limits are plotted as dashed lines using column 3 as a single X column versus columns 7 and 8 as many Y columns, and the prediction limits are plotted as dotted lines using column 3 as a single X column versus columns 7 and 8 as many Y columns. 7. To create your own graph in SigmaPlot, create a Scatter Plot with a Simple Regression, plotting column 1 against column 2 as the symbols and using column 3 plotted against column 4 as the regression. Add confidence and prediction intervals using column 3 as the X column and columns 7 and 8 as the Y columns.
652 Chapter 13
3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the LOWPFILT.XFM transform file in the XFMS directory. The Low Pass Filter transform appears in the edit window. 4. Set the sampling interval dt (the time interval between data points) and the half power point fc values. The half power point is the frequency at which the squared magnitude of the frequency response is reduced by half of its magnitude at zero frequency. 5. If necessary, change the cy1 source column value and cy2 filtered data results to the correct column numbers. 6. Click Run to run the transform. Filtered data appears in column 3 in the worksheet, or in the worksheet column you specified in the transform. 7. If you opened the Low Pass Filter graph, view the graph page. The second graph appears as a line graph plotting the smoothed data in columns 1 and 3. 8. To create your own graphs in SigmaPlot, create the first graph as a Line Plot with a Simple Spline Curve using the raw data in columns 1 and 2 as the X and Y data. Make the second Line Plot graph with a Simple Spline Curve using the data in column1 as the X data and the smoothed data in column 3 as the Y data.
Figure 13-40 Low Pass Filter Graph Plotting Raw Data and Filtered Data
Lowess Smoothing
Smoothing is used to elicit trends from noisy data. Lowess smoothing produces smooth curves under a variety of conditions. "Lowess" means locally weighted regression. Each point along the smooth curve is obtained from a regression of data points close to the curve point with the closest points more heavily weighted. The y value of the data point is replaced by the y value on the regression line. The amount of smoothing, which affects the number of points in the regression, is specified by the user with the parameter f. This parameter is the fraction of the total number of points that is used in each regression. If there are 50 points along the smooth curve with f = 0.2 then 50 weighted regressions are performed and each regression is performed using 10 points. An example of the use of lowess smoothing for the U.S. wheat production from 1872 to 1958 is shown in the figures below. The smoothing parameter f was chosen to be 0.2 since this produced a good tradeoff between noisy undersmoothing and oversmoothing which misses some of the peak-and-valley details in the data. 1. To use the provided sample data and graph, open the Lowess Smoothing worksheet and graph in the Lowess Smoothing section of the Transform Examples notebook. The worksheet appears with data in columns 1, 2, and 3. 2. To use your own data, enter the XY data for your curve in columns 1 and 2, respectively. If your data has been placed in other columns, you can specify these columns after you open the LOWESS.XFM transform file. Enter data into an existing or a new worksheet. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the LOWESS.XFM transform file in the Transforms directory. The Lowess transform appears in the edit window. 4. Click Run. The results are placed in column 3 of the worksheet, or in the column specified by the ouput variable. 5. If you opened the Lowess Smoothing graph, view the graph page. The smoothed curve is plotted on the second graph and both the original and smoothed data are plotted on the third.
654 Chapter 13
Figure 13-41 U.S. Wheat data and the lowess smoothed curve (f = 0.2). Notice the definite decreased production during World War II.
If you want to plot your own results, create a line plot of column 1 versus column 3.
Normalized Histogram
This simple transform creates a histogram normalized to unit area. The resulting data can be graphed as a bar chart. Histogram bar locations are shifted to be placed over the histogram box locations. The resulting bar chart is an approximation to a probability density function. To calculate and graph a normalized histogram sample, you can either use the provided sample data and graph or begin a new notebook, enter your own data, and create your own graph using the data. 1. To use the provided sample data and graph, open the Normalized Histogram worksheet and graph in the Normalized Histogram and Graph section of the Transform Examples
notebook. The worksheet appears with data in column 1. The data is made up of exponentially distributed random numbers generated with the transform: x = random(200,1,1.e-10,1)col(1) = -ln(x) The graph page appears with an empty graph. 2. To use your own data, place your data in column 1 of the worksheet. If your data has been placed in another column, you can specify this column after you open the NORMHIST.XFM transform file. You can enter data into an existing or new worksheet. 3. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the NORMHIST.XFM transform file in the XFMS directory. The Normalized Histogram transform appears in the edit window. 4. Click Run. The results are placed in columns 2 and 3 of the worksheet, or in the columns specified by the res variable. 5. If you opened the Normalized Histogram graph, view the graph page. A histogram appears using column 2 as X data versus column 3 as the Y data. 6. To create your own graph in SigmaPlot, create a Vertical Bar chart with simple bars, then set the bar widths as wide as possible.
Figure 13-42 Normalized Histogram Graph
656 Chapter 13
5. If you opened the sample Smooth Color Transition graph, view the graph page. A Scatter Plot appears plotting column 2 as a Simple Scatter plot style using Single Y data format. The symbol colors are obtained by specifying column 3 in the Symbols, Fill Color drop-down list in the Plots panel of the Graph Properties dialog box. The Smooth Color Transition transform applies gradually changing colors to each of the data points. The smaller residual values are colored blue, which gradually changes to red for the larger residuals. 6. To create your own graph using SigmaPlot, make a Scatter Plot graph with a Scatter Plot with Simple Scatter style. Plot the data as Single Y data format. Use the color cells produced by the transform by selecting the corresponding worksheet column from the Symbol Fill Color drop-down list.
658 Chapter 13
6. Check for any ties between true response and censored data. If any exist, make sure that within the tied data, the censored data follows the true response data. 7. Click Run to run the file. The sorted time, cumulative survival probability, and the standard error are placed in columns res, res+1, and res+2, respectively. For graphical purposes a zero, one, and zero have been placed in the first rows of the sorted time, cumulative survival curve probability and standard error columns. 8. If you opened the sample Survival graph, view the page. The Simple Horizontal Step Plot graphs the survival curve data from columns res as the X data versus column res+1 as the Y data and a Scatter Plot graphs the data from the same columns. The first data point of the Scatter Plot at (0,1) is not displayed by selecting rows 2 to end in the Portions of Columns Plotted area of the Data section in the Plots tab of the Graph Properties dialog box. As shown in the figure below, a tied censored data point has been incorrectly placed; it should follow uncensored data. 9. To graph a survival curve using SigmaPlot, create a Line graph with a Simple Horizontal Step Plot graphing column res as the X data versus column res+1 as the Y data. If desired, create an additional Scatter plot, superimposing the survival data using the same columns for X data and Y data. To turn off the symbol drawn at x = 0 and y = 1, select Plot 2 and set Only rows = 2 to end by 1 in the Plots tab and Data sections of the Graph Properties dialog box.
660 Chapter 13
1. To use the sample worksheet and graph, double-click the graph page icon in the User Defined Axis Scale section of the Transforms Examples notebook. The User Defined Axis Scale worksheet appears with data in columns 1 through 3. The graph page appears with an empty graph with gridlines. 2. To use your own data, place your original X data in column 1, Y data in column 2, and the Y axis tick interval values in column 3. If your data has been placed in other columns, you can specify these columns after you open the USERAXIS.XFM file. 3. Press F10 to open the User-Defined Transform dialog box, then open the USERAXIS.XFM transform. If necessary, change the y_col, tick_col, and res variables to the correct column numbers. 4. Click Run. The results are placed in columns 4 and 5, or the columns specified by the res variable. 5. If you opened the User Defined Axis Scale graph, view the page. The graph is already set up to plot the data and grid lines. 6. To plot the transformed Y data using SigmaPlot, plot column 1 as the X values versus column 4 as the Y values. To plot the Y axis tick marks, open the Ticks panel under the Axes tab of the Graph Properties dialog box. Select Column 5 from the Major Tick Intervals drop-down list. To draw the tick labels, use the Y tick interval data as the tick label source by selecting Column 3 from the Tick Label Type drop-down list in the Tick Labels panel under the Axes tab of the Graph Properties dialog box.
Vector Plot
The VECTOR.XFM transform creates a field of vectors (lines with arrow heads) from data which specifies the X and Y position, length, and angle of each vector. The data is entered into four columns. Executing the transform produces six columns of three XY pairs, which describe the arrow body and the upper and lower components of the arrow head. Other settings are: The length of the arrow head. The angle in degrees between the arrow head and the arrow body. The length of the vector (if you want to specify it as a constant). To generate a vector plot, you can either use the provided sample data and graph or begin a new notebook, enter your own data, and create your own graph using the data. 1. To use the sample worksheet and graph, double-click the graph page icon in the Vector section of the Transform Examples notebook. The Vector worksheet appears with data in columns 1 through 4. The graph page appears with an empty graph.
662 Chapter 13
2. To use your own data, enter the vector information into the worksheet. Data must be entered in four column format, with the XY position of the vector starting in the first column, the length of the vectors (which correspond to the axis units), and the angle of the vector, in degrees. The default starting column for this block is column one. 3. Press F10 to open the User-Defined Transforms dialog box, then click the Open button to open the VECTOR.XFM file in the XFMS directory. 4. If necessary, change the starting worksheet column for your vector data block xc. 5. If desired, change the default arrowhead length L (in axis units) and the Angle used by the arrowhead lines. This is the angle between the main line and each arrowhead line. 6. If you want to use vectors of constant length, set the l value to the desired length, then uncomment the remaining two lines under the Constant Vector Length heading. 7. Make sure that Radians are selected as the Trigonometric Units (they should be by default. 8. Click Run to run the transform. The transform produces six columns of three XY pairs, which describe the arrow body and the upper and lower components of the arrow head. 9. If you opened the Vector graph, view the page. The Line Plot with Multiple Straight Line appears plotting columns 5 through 10 as XY pairs. 10. To plot the vector data using SigmaPlot, create a Line Plot with Multiple Straight Line graph that plots columns 5 through 10 as three vector XY column pairs.
664 Chapter 13
curve data plotted over some sample root locus data. This plot uses columns 1 and 2 as the first curve and columns 3 and 4 as the second curve. 2. To use your own data, place your root locus, zero, and pole data in columns 1 through 10. If your locus data has been placed in other columns, you can change the location of the results columns after you open the ZPLANE.XFM file. 3. To plot the design curves of your data, create a Line Plot with Multiple Spline Curves, then plot column 1 as the X data against column 2 as the Y data for the first curve and column 3 as the X data against column 4 as the Y data as the second curve. 4. Press F10 to open the User-Defined Transform dialog box, then click the Open button, and open the ZPLANE.XFM transform in the XFMS directory. If necessary, change the res variable to the correct column number. 5. Click Run. The results are placed in columns 11 through 20, or the columns specified by the res variable. 6. If you opened the Z Plane graph, view the page. The circle, frequency trajectory, and damping trajectory data is automatically plotted with the design data. 7. To plot the circle data using SigmaPlot, create Multiple Line Plots with Simple Spline Curves. For the first plot use column 11 as the X values versus column 12 as the Y values. 8. To plot the frequency trajectory data (zeta) plot column 13 versus column 14 and column 15 versus column 16 as the XY pairs. 9. To plot the damping trajectory data (omega) plot column 17 versus column 18 and column 19 versus column 20 as the XY pairs.
666 Chapter 13
Chapter
Nonlinear Regression
14
What is Regression?
Regression is most often used by scientists and engineers to visualize and plot the curve that best describes the shape and behavior of their data. Regression procedures find an association between independent and dependent variables that, when graphed on a Cartesian coordinate system, produces a straight line, plane or curve. This is also commonly known as curve fitting. The independent variables are the known, or predictor, variables. These are most often your X-axis values. When the independent variables are varied, they result in corresponding values for the dependent, or response, variables, most often assigned to the Y-axis. Regression finds the equation that most closely describes, or fits, the actual data, using the values of one or more independent variables to predict the value of a dependent variable. The resulting equation can then be plotted over the original data to produce a curve that fits the data.
667
668 Chapter 14
graphically presented and organized by different categories, making selection of your models straightforward. Built-in shortcuts let you bypass all but the simplest procedures; fitting a curve to your data can be as simple as picking the equation to use, then clicking a button. Note: For more complicated curve fitting, try using the Dynamic Fit Wizard. For more information, see Dynamic Curve Fitting below. Use the Regression Wizard to: Select the function describing the shape of your data. SigmaPlot provides over 100 built-in equations. You can also create your own custom regression equations.For more information, see Regression Equation Library in Chapter 18. Select the variables to fit to the function. You can select your variables from either a graph or a worksheet. Evaluate and save your results. You can automatically plot the resulting curves on a graph, and save statistical results to the worksheet and text reports. The Regression Wizard is also compatible with older .FIT files. For more information, see Opening .FIT Files below.
Curve-fitting Algorithm
The SigmaPlot curve fitter uses the Marquardt-Levenberg algorithm to find the coefficients (parameters) of the independent variable(s) that give the best fit between the equation and the data.
This algorithm seeks the values of the parameters that minimize the sum of the squared differences between the values of the observed and predicted values of the dependent variable.
is the predicted value of the dependent variable. This process is iterativethe curve fitter begins with a guess at the parameters, checks to see how well the equation fits, then continues to make better guesses until the differences between the residual sum of squares no longer decreases significantly. This condition is known as convergence. For more information, see References for the Marquardt-Levenberg Algorithm below.
670 Chapter 14
Shrager, R.I. (1972). Quadratic Programming for N. Communications of the ACM, 15, 41-45.
Select SigmaPlot Curve Fit as the file type. .fit files are opened as a single equation in a notebook. You can also open .fit files from the Library panel of both the Regression Wizard and the Dynamic Fit Wizard. For more information, see Dynamic Curve Fitting below.
1. View the page or worksheet with the data you want to fit.
If you select a graph, right-click the curve you want fitted, and on the shortcut menu, click Fit Curve. Note: If you are running a regression from the graph page, make sure you select the plot itself, not the graph, or Fit Curve will not appear on the shortcut menu. If you are using a worksheet, select the variables in the worksheet you want to fit, then from the menus select:
Statistics Nonlinear Regression Regression Wizard
672 Chapter 14
Selecting the variable from the Variable Columns drop-down list in the Regression Wizard. The equation picture to the left prompts you for which variable to select. You can also modify other equation settings and options from this panel by clicking Options, which opens the Equations Options dialog box. These options include changing initial parameter estimates, parameter constraints, weighting, and other related settings. For more information, see Equation Options below. If you pick variables from a worksheet column, you can also set the data format. For more information, see Variable Options below. 4. When you have selected your variables, you can either click Finish, or click Next to view the Initial Results.
674 Chapter 14
Tip: Remember, these settings are reused each time you perform a regression and overwrite data if it exists in these columns in subsequent worksheets. To avoid overwriting data, use First Empty to place the fitted curve results in empty columns. 18. Click Finish.
Assumption Checking
Select the Assumption Checking tab from the Report Options for Nonlinear Regression to view the Normality, Constant Variance, and Durbin-Watson options. These options test your data for its suitability for regression analysis by checking three assumptions that a linear regression makes about the data. A nonlinear regression assumes: That the source population is normally distributed about the regression. The variance of the dependent variable in the source population is constant regardless of the value of the independent variable(s). That the residuals are independent of each other. All assumption checking options are selected by default. Only disable these options if you are certain that the data was sampled from normal populations with constant variance and that the residuals are independent of each other. Normality Testing. SigmaPlot uses the Kolmogorov-Smirnov test to test for a normally distributed population. Constant Variance Testing. SigmaPlot tests for constant variance by computing the Spearman rank correlation between the absolute values of the residuals and the observed value of the dependent variable. When this correlation is significant, the constant variance assumption may be violated, and you should consider trying a different model (i.e., one that more closely follows the pattern of the data), or transforming one or more of the independent variables to stabilize the variance. For more information, see User-Defined Transforms in Chapter 13. P Values for Normality and Constant Variance. The P value determines the probability of being incorrect in concluding that the data is not normally distributed (P value is the risk of falsely rejecting the null hypothesis that the data is normally distributed). If the P computed by the test is greater than the P set here, the test passes. To require a stricter adherence to normality and/or constant variance, increase the P value. Because the parametric statistical methods are relatively robust in terms of detecting violations of the assumptions, the suggested value in SigmaPlot is 0.05. Larger values of P (for example, 0.10) require less evidence to conclude that the residuals are not normally distributed or the constant variance assumption is violated. To relax the requirement of normality and/or constant variance, decrease P. Requiring smaller values of P to reject the normality assumption means that you are willing to
676 Chapter 14
accept greater deviations from the theoretical normal distribution before you flag the data as non-normal. For example, a P value of 0.01 for the normality test requires greater deviations from normality to flag the data as non-normal than a value of 0.05. Note: Although the assumption tests are robust in detecting data from populations that are non-normal or with non-constant variances, there are extreme conditions of data distribution that these tests cannot detect. However, these conditions should be easily detected by visually examining the data without resorting to the automatic assumption tests. Durbin-Watson Statistic. SigmaPlot uses the Durbin-Watson statistic to test residuals for their independence of each other. The Durbin-Watson statistic is a measure of serial correlation between the residuals. The residuals are often correlated when the independent variable is time, and the deviation between the observation and the regression line at one time are related to the deviation at the previous time. If the residuals are not correlated, the Durbin-Watson statistic will be 2. Difference from 2 Value. Enter the acceptable deviation from 2.0 that you consider as evidence of a serial correlation in the Difference for 2.0 box. If the computed DurbinWatson statistic deviates from 2.0 more than the entered value, SigmaPlot warns you that the residuals may not be independent. The suggested deviation value is 0.50, i.e., Durbin-Watson Statistic values greater than 2.5 or less than 1.5 flag the residuals as correlated. To require a stricter adherence to independence, decrease the acceptable difference from 2.0. To relax the requirement of independence, increase the acceptable difference from 2.0.
Residuals
Click the Residuals tab in the Report Options for Nonlinear Regression dialog box to view the Predicted Values, Raw, Standardized, Studentized, Studentized Deleted, and Report Flagged Values Only options. Studentized Residuals. Studentized residuals scale the standardized residuals by taking into account the greater precision of the regression line near the middle of the data versus the extremes. The Studentized residuals tend to be distributed according to the Student t distribution, so the t distribution can be used to define "large" values of the Studentized residuals. SigmaPlot automatically flags data points with "large" values of
the Studentized residuals, i.e., outlying data points; the suggested data points flagged lie outside the 95% confidence interval for the regression population. To include studentized residuals in the report, make sure this check box is selected. Click the selected check box if you do not want to include studentized residuals in the worksheet. Studentized Deleted Residuals. Studentized deleted residuals are similar to the Studentized residual, except that the residual values are obtained by computing the regression equation without using the data point in question. To include Studentized deleted residuals in the report, make sure this check box is selected. Click the selected check box if you do not want to include studentized deleted residuals in the worksheet. SigmaPlot can automatically flag data points with "large" values of the studentized deleted residual, i.e., outlying data points; the suggested data points flagged lie outside the 95% confidence interval for the regression population. Note: Both Studentized and Studentized deleted residuals use the same confidence interval setting to determine outlying points. Raw Residuals. The raw residuals are the differences between the predicted and observed values of the dependent variables. To include raw residuals in the report, make sure this check box is selected. Click the selected check box if you do not want to include raw residuals in the worksheet. To assign the raw residuals to a worksheet column, select the number of the desired column from the corresponding drop-down list. If you select none from the drop-down list and the Raw check box is selected, the values appear in the report but are not assigned to the worksheet. Predicted Values. Use this option to calculate the predicted value of the dependent variable for each observed value of the independent variable(s), then save the results to the worksheet. Click the selected check box if you do not want to include raw residuals in the worksheet. To assign predicted values to a worksheet column, select the worksheet column you want to save the predicted values, from the corresponding drop-down list. If you select none and the Predicted Values check box is selected, the values appear in the report but are not assigned to the worksheet. Standardized Residuals. The standardized residual is the residual divided by the standard error of the estimate. The standard error of the residuals is essentially the standard deviation of the residuals, and is a measure of variability around the
678 Chapter 14
regression line. To include standardized residuals in the report, make sure this check box is selected. Click the selected check box if you do not want to include raw residuals in the worksheet. Flag Values >.SigmaPlot automatically flags data points lying outside of the confidence interval specified in the corresponding box. These data points are considered to have "large" standardized residuals, i.e., outlying data points. You can change which data points are flagged by editing the value in the Flag Values > edit box. The suggested residual value is 2.5. Report Flagged Values Only. To include only the flagged standardized and Studentized deleted residuals in the report, make sure the Report Flagged Values Only check box is selected. Clear this option to include all standardized and Studentized residuals in the report.
More Statistics
Click the More Statistics tab in the Report Options for Nonlinear Regression dialog box to view options for Confidence and Prediction Intervals and PRESS Prediction Error. Confidence Intervals. You can set the confidence interval for the population, regression, or both and then save them to the worksheet. Prediction Interval. The confidence interval for the population gives the range of values that define the region that contains the population from which the observations were drawn. To include confidence intervals for the population in the report, make sure the Population check box is selected. Click the selected check box if you do not want to include the confidence intervals for the population in the report. Confidence Interval. The confidence interval for the regression line gives the range of values that defines the region containing the true mean relationship between the dependent and independent variables, with the specified level of confidence. To include confidence intervals for the regression in the report, make sure the Regression check box is selected, then specify a confidence level by entering a value in the percentage box. The confidence level can be any value from 1 to 99. The suggested confidence level for all intervals is 95%.
Click the selected check box if you do not want to include the confidence intervals for the population in the report. Click the selected check box if you do not want to include the confidence intervals for the population in the report. Saving Confidence Intervals to the Worksheet. To save the confidence intervals to the worksheet, select the column number of the first column you want to save the intervals to from the Starting in Column drop-down list. The selected intervals are saved to the worksheet starting with the specified column and continuing with successive columns in the worksheet. PRESS Prediction Error. The PRESS Prediction Error is a measure of how well the regression equation fits the data. Leave this check box selected to evaluate the fit of the equation using the PRESS statistic. Click the selected check box if you do not want to include the PRESS statistic in the report.
Other Diagnostics
Click the Other Diagnostics tab in the Report Options for Nonlinear Regression dialog box to view options Influence, DFFITS, leverage, Cooks Distance and power. Influence. Influence options automatically detect instances of influential data points. Most influential points are data points which are outliers, that is, they do not do not "line up" with the rest of the data points. These points can have a potentially disproportionately strong influence on the calculation of the regression line. You can use several influence tests to identify and quantify influential points. DFFITS. DFFITS is the number of estimated standard errors that the predicted value changes for the ith data point when it is removed from the data set. It is another measure of the influence of a data point on the prediction used to compute the regression coefficients. Predicted values that change by more than two standard errors when the data point is removed are considered to be influential. Select DFFITS to compute this value for all points and flag influential points, i.e. those with DFFITS greater than the value specified in the Flag Values > edit box. The suggested value is 2.0 standard errors, which indicates that the point has a strong influence on the data. To avoid flagging more influential points, increase this value; to flag less influential points, decrease this value. Leverage. Leverage is used to identify the potential influence of a point on the results of the regression equation. Leverage depends only on the value of the
680 Chapter 14
independent variable(s). Observations with high leverage tend to be at the extremes of the independent variables, where small changes in the independent variables can have large effects on the predicted values of the dependent variable. Select Leverage to compute the leverage for each point and automatically flag potentially influential points, i.e., those points that could have leverages greater than the specified value times of the expected leverage. The suggested value is 2.0 times the expected leverage for the regression. To avoid flagging more potentially influential points, increase this value; to flag points with less potential influence, lower this value. Cooks Distance. Cooks distance is a measure of how great an effect each point has on the estimates of the parameters in the regression equation. Cooks distance assesses how much the values of the regression coefficients change if a point is deleted from the analysis. Cooks distance depends on both the values of the independent and dependent variables. Select Cooks Distance to compute this value for all points and flag influential points, i.e., those with a Cooks distance greater than the specified value. The suggested value is 4.0. Cooks distances above 1 indicate that a point is possibly influential. Cooks distances exceeding 4 indicate that the point has a major effect on the values of the parameter estimates. To avoid flagging more influential points, increase this value: to flag less influential points, lower this value. Power. The power of a regression is the power to detect the observed relationship in the data. The alpha is the acceptable probability of incorrectly concluding there is a relationship. Select Power to compute the power for the linear regression data. Change the alpha value by editing the number in the Alpha Value edit box. The suggested value is = 0.05. This indicates that a one in twenty chance of error is acceptable, or that you are willing to conclude there is a significant relationship, when P < 0.05. Report Flagged Values Only. To only include only the influential points flagged by the influential point tests in the report, select Report Flagged Values Only. Clear this option to include all influential points in the report.
convenient if you have created or stored equations along with the rest of your graphs and data. 1. In the Notebook Manager, view the notebook with the equation you want to use, and double-click the equation. The Regression Wizard appears with the equation selected. 2. Select the variables as prompted by clicking a curve or worksheet columns. Note that at this point you can open and view any notebook, worksheet or page you would like, and pick your variables from that source. 3. Click Finish to complete the regression, or click Next if you want to view initial results or change your results options.
682 Chapter 14
Add As, add the equation to the desired section, and then edit the Equations, Variables and Parameters as desired. You can also copy and paste equations from notebook to notebook like any other notebook item. You can also edit pasted built-in equations. For more information, see Editing Code in Chapter 15.
Variable Options
Data Format Options. If you use data columns from the worksheet, you can specify the data format to use in the variables panel of the Regression Wizard. By default, the data format when assigning columns from the worksheet is XY Pair. The data format options are: XY pair. Select an x and a y variable. Y only. Select only a y variable column. XY column means. Pick one x column, then multiple y columns; the y columns will be graphed as means. Y column means only. Pick multiple y columns; the columns will be graphed as means. From Code. Uses the current settings as shown when editing code. XY Replicate. Select X and multiple Y columns. Rows of the Y columns are replicate measurements.
Y Replicates. Select multiple Y columns. Rows of the Y column are replicate measurements. When you use an existing graph as your data source, the Regression Wizard displays a format reflecting the data format of the graph. You cannot change this format unless you switch to using the worksheet as your data source, or run the regression directly from editing the code.
Equation Options
If the curve fitter fails to find a good fit for the curve, you can try changing the equation options to see if you can improve the fit. To set options for a regression, click the Options button in the Variables panel of the Regression, Dynamic Fit, or Global Fit Wizard. The Equation Options dialog box appears. Note: If you want to edit the settings in the equation document manually, click the Edit Code button. For more information, see Editing Code in Chapter 15. Use the Equation Options dialog box to: Change initial parameter values. Add or change constraints. Change constant values. Use weighted fitting, if it is available. Change convergence options.
Parameters
The default setting for the initial parameter value is shown Automatic. The Automatic setting available with the built-in SigmaPlot equations uses algorithms that analyze
684 Chapter 14
your data to predict initial parameter estimates. These do not work in all cases, so you may need to enter a different value. Just click the parameter you want to change, and make the change in the edit box. The values that appear in the Initial Parameters drop-down list were previously entered as parameter values. Any parameter values you enter will also be retained between sessions.
Figure 14-1 Selecting Numeric Initial Parameters in the Equation Options dialog box.
Parameters can be either a numeric value or a function. The value of the parameter should approximate the final result, in order to help the curve fitter reach a valid result, but this depends on the complexity and number of parameters of the equation. Often an initial parameter nowhere near the final result will still work. However, a good initial estimate helps guarantee better and faster results.
Constraints
Use constraints to set limits and conditions for parameter values, restricting the regression search range and improving curve fitter speed and accuracy. Constraints are often unnecessary, but should always be used whenever appropriate for your model.
Constraints are also useful to prevent the curve fitter from testing unrealistic parameter values. For example, if you know that a parameter should always be negative, you can enter a constraint defining the parameter to be always less than 0. You can also use constraints if the regression produces parameter values that you know are inaccurate. Simply click Back from the initial results panel, click the Options button, and enter constraint(s) that prevent the wrong parameter results. Note that a parameter equals a constraint value at the completion of the fit, the constraint is called active. You can view these constraints from the initial results panel by clicking View Constraints. For more information, see Checking Use of Constraints below.
686 Chapter 14
Defining Constants
Constants that appear in the Constants edit window have been previously defined as a constant, rather than a parameter to be determined by the regression. To edit a constant value, or define new constant values, click Edit Code on the Regression Wizard, Dynamic Fit Wizard or Global Fit Wizarddialog box. Constants are defined when an equation is created. Currently, you can only define new constants by editing the regression equation code. However, you can redefine any existing constants. Change only the value of the constant. Do not add new constant values; constant variables must exist in the equation and not defined already under variables or parameters, so they can only be defined within the code of an equation.
Iterations
The Iterations option sets the maximum number of repeated fit attempts before failure. Each iteration of the curve fitter is an attempt to find the parameters that best fit the model. With each iteration, the curve fitter varies the parameter values incrementally, and tests the fit of that model to your data. When the improvement in the fit from one iteration to the next is smaller than the setting determined by the Tolerance option, the curve fitter stops and displays the results.
Figure 14-3 Changing iterations
Change the number of iterations to speed up or improve the regression process, especially if a complex fit requires more than the default of 100 iterations. You can also reduce the number of iterations if you want to end a fit to check on its interim progress before it takes too many iterations. Setting Iterations to 0 will simply evaluate the dependent variable of the fit equation using the initial parameter values. To change the maximum number of iterations, enter the number of iterations to use, or select a previously used number of iterations from the drop-down list. When the maximum number of iterations is reached, the regression stops and the current results are displayed in the initial parameters panel. If you want to continue with more iterations, you can click More Iterations on the Regression Wizard.
688 Chapter 14
Step Size
Step size, or the limit of the initial change in parameter values used by the curve fitter as it tries, or iterates, different parameter values, is a setting that can be changed to speed up or improve the regression process.
Figure 14-4 Changing Step Size
A large step size can cause the curve fitter to wander too far away from the best parameter values, whereas a step size that is too small will result in slow convergence to the best parameters. For most functions, the default step size value is 1. To change the Step Size value, type the desired step size in the Step Size edit box, or select a previously defined value from the drop-down list.
Tolerance
The Tolerance option controls the condition that must be met in order to end the regression process. When the absolute value of the difference between the sum of squares of the residuals (square root of the sum of squares of the residuals), from one iteration to the next, is less than the tolerance value, the iteration stops.
When the tolerance condition has been met, a minimum of the sum of squares has usually been found, which indicates a correct solution. However, local minima in the sum of squares can also cause the curve fitter to find an incorrect solution. Decreasing the value of the tolerance makes the requirement for finding an acceptable solution more strict; increasing the tolerance relaxes this requirement. The default tolerance setting is1e-10. To change the tolerance value, type the desired value in the Tolerance edit box, or select a previously defined value from the drop-down list.
690 Chapter 14
Cancelling a Regression
To stop a regression while it is running, click Cancel. The initial results appear, displaying the most recent parameter values, and the sum of square value. You can continue the regression process by clicking More Iterations on the Regression Wizard.
Rsqr
R2 is the coefficient of determination, the most common measure of how well a regression model describes the data. The closer R2 is to one, the better the independent variables predict the dependent variable. R2 equals 0 when the values of the independent variable does not allow any prediction of the dependent variables, and equals 1 when you can perfectly predict the dependent variables from the independent variables.
Fit Results
The initial results are displayed in the results window, in five columns. Parameter. The parameter names are shown in the first column. These parameters are derived from the original equation. Value. The calculated parameter values are shown in the second column. StdErr. The asymptotic standard errors of the parameters are displayed in column three. The standard errors and coefficients of variation can be used as a gauge of the fitted curves accuracy. CV(%). The parameter coefficients of variation, expressed as a percentage, are displayed in column four. This is the normalized version of the standard errors:
The coefficient of variation values and standard errors can be used as a gauge of the accuracy of the fitted curve. Dependency. The last column shows the parameter dependencies. The dependence of a parameter is defined to be
Parameters with dependencies near 1 are strongly dependent on one another. This may indicate that the equation(s) used are too complicated and over-parameterizedtoo many parameters are being used, and using a model with fewer parameters may be better.
692 Chapter 14
More Iterations
If the maximum number of iterations was reached before convergence, or if you canceled the regression, the More Iterations button is available. Click More Iterations to continue for as many iterations as specified by the Iterations option, or until completion of the regression.
The Constraints dialog box displays all constraints, and flags the ones encountered with the word (active). A constraint is flagged as active when the parameter values lie on the constraint boundary. For example, the constraint: a+b<1 is active when the parameters satisfy the condition a+b=1, but if a+b<1, the constraint is inactive. Note that an equality constraint is always active (unless there are constraint inconsistencies).
694 Chapter 14
Figure 14-8 Generating and Saving a Report from the Regression Wizard
Create Report. Select to save regression reports to the current notebook section. Adding Equation to Notebook. To add the current regression equation to the current notebook, select Add Equation to Notebook. If this option is selected, a copy of the equation is added to the current section of your notebook.
To create a new graph, select Create New Graph. Click Finish to create a new notebook section containing a worksheet of the plotted data and graph page.
696 Chapter 14
From this panel you can select worksheet columns for X, Y, (and Z data for 3D graphs) by clicking worksheet columns. The default of First Empty places the results in the first available column after the last filled cell.
where gradF is the (parameter) gradient of the model F, evaluated at x and at the bestfit parameter values, and Cov is the covariance matrix computed at the final iteration of the regression. After computing c, the upper and lower limits of both bands are given by:
698 Chapter 14
where
is the predicted value at x. MSres is the mean residual sum of squares. conf.level% is the percent confidence level. DF is the residual degree of freedom (number of data points - number of parameters). CriticalT(conf.level%,DF) is the inverse T-Distribution with DF degrees of freedom evaluated 1-/2, where satisfies conf.level% = 100(1-)%
While the default confidence level for confidence and prediction bands is 95%, you can change this value in the Report Options for Nonlinear Regression dialog box. 1. Click Back if you are viewing the Regression Wizard - Graph Options dialog box to the Numeric Output Options panel. For more information, see Using the Regression Wizard above.
Figure 14-13 Click Report to open the Options for Nonlinear Regression dialog box.
2. Click Report to open the Options for Nonlinear Regressions dialog box. Here is where you control what you would like to appear in the report.
Figure 14-14 Selecting the Report Options
3. Click the More Statistics tab. 4. Under Confidence Intervals, select the Prediction and Confidence intervals you would like to appear in the report. 5. Set the percentage for the confidence level in the Confidence Level box.
700 Chapter 14
6. In the Starting in Column drop-down list, select First Available Column or select (None) if you do not want the confidence and prediction intervals to appear in the worksheet, 7. Click OK to save the changes and close the dialog box. 8. Click Next in the Regression Wizard. The Regression Wizard - Graph Options panel appears with the new percentage for the confidence and prediction bands.
Figure 14-15 Selecting Confidence and Prediction Bands
When you click Finish in the Regression Wizard, the confidence and prediction bands appear on the graph. If you selected First Available Column in the More Statistics tab of the Report Options for Nonlinear Regression dialog box, beginning in the first empty column, four columns of graph data appear in the worksheet which represent the upper and lower limits of the confidence and prediction bands.
Figure 14-16 Worksheet Columns Representing Data for the Confidence and Prediction Bands
These results also appear in the report. For more information, see Confidence and Prediction Intervals below.
Equation Code
This is a printout of the code used to generate the regression results. For more information, see Editing Code in Chapter 15.
702 Chapter 14
R and R Squared
The multiple correlation coefficient, R, and the coefficient of determination, R2, are both measures of how well the regression model describes the data. R values near 1 indicate that the equation is a good description of the relation between the independent and dependent variables. R equals 0 when the values of the independent variable does not allow any prediction of the dependent variables, and equals 1 when you can perfectly predict the dependent variables from the independent variables.
Adjusted R Squared
The adjusted R2, Radj2, is also a measure of how well the regression model describes the data, but takes into account the number of independent variables, which reflects the degrees of freedom. Larger Radj2 values (nearer to 1) indicate that the equation is a good description of the relation between the independent and dependent variables.
You can conclude from large t values that the independent variable can be used to predict the dependent variable (for example., that the coefficient is not zero). P value. P is the P value calculated for t. The P value is the probability of being wrong in concluding that the coefficient is not zero (i.e., the probability of falsely rejecting the null hypothesis, or committing a Type I error, based on t). The smaller the P value, the greater the probability that the coefficient is not zero. Traditionally, you can conclude that the independent variable can be used to predict the dependent variable when P < 0.05.
704 Chapter 14
The residual mean square is a measure of the variation of the residuals about the regression plane, or
F statistic
The F test statistic gauges the contribution of the independent variables in predicting the dependent variable. It is the ratio
If F is a large number, you can conclude that the independent variables contribute to the prediction of the dependent variable (i.e., at least one of the coefficients is different from zero, and the unexplained variability is smaller than what is expected from random sampling variability of the dependent variable about its mean). If the F ratio is around 1, you can conclude that there is no association between the variables (i.e., the data is consistent with the null hypothesis that all the samples are just randomly distributed). P value. The P value is the probability of being wrong in concluding that there is an association between the dependent and independent variables (i.e., the probability of falsely rejecting the null hypothesis, or committing a Type I error, based on F ). The smaller the P value, the greater the probability that there is an association. Traditionally, you can conclude that the independent variable can be used to predict the dependent variable when P < 0.05.
PRESS Statistic
PRESS, the Predicted Residual Error Sum of Squares, is a gauge of how well a regression model predicts new data. The smaller the PRESS statistic, the better the predictive ability of the model. The PRESS statistic is computed by summing the squares of the prediction errors (the differences between predicted and observed values) for each observation, with that point deleted from the computation of the regression equation.
Durbin-Watson Statistic
The Durbin-Watson statistic is a measure of correlation between the residuals. If the residuals are not correlated, the Durbin-Watson statistic will be 2; the more this value differs from 2, the greater the likelihood that the residuals are correlated.
706 Chapter 14
Regression assumes that the residuals are independent of each other; the DurbinWatson test is used to check this assumption. If the Durbin-Watson value deviates from 2 by more than 0.50, a warning appears in the report, i.e., if the Durbin-Watson statistic is below 1.50 or above 2.50.
Normality Test
The normality test results display whether the data passed or failed the test of the assumption that the source population is normally distributed around the regression, and the P value calculated by the test. All regressions assume a source population to be normally distributed about the regression line. If the normality test fails, a warning appears in the report. Failure of the normality test can indicate the presence of outlying influential points or an incorrect regression model.
Figure 14-18 Regression Report Showing Normality Test Results
regression, or transforming the independent variable to stabilize the variance and obtain more accurate estimates of the parameters in the regression equation. If you perform a weighted regression, the normality and equal variance tests use the weighted residuals
Power
The power, or sensitivity, of a regression is the probability that the model correctly describes the relationship of the variables, if there is a relationship. Regression power is affected by the number of observations, the chance of erroneously reporting a difference a (alpha), and the slope of the regression. Alpha. Alpha (a) is the acceptable probability of incorrectly concluding that the model is correct. An a error is also called a Type I error (a Type I error is when you reject the hypothesis of no association when this hypothesis is true). Smaller values of a result in stricter requirements before concluding the model is correct, but a greater possibility of concluding the model is incorrect when it is really correct (a Type II error). Larger values of a make it easier to conclude that the model is correct, but also increase the risk of accepting an incorrect model (a Type I error).
Regression Diagnostics
The regression diagnostic results display the values for the predicted values, residuals, and other diagnostic results. Row. This is the row number of the observation. Predicted Values. This is the value for the dependent variable predicted by the regression model for each observation. Residuals. These are the unweighted raw residuals, the difference between the observed and predicted values for the dependent variables.
708 Chapter 14
Standardized Residuals. The standardized residual is the raw residual divided by the standard error of the estimate. If the residuals are normally distributed about the regression, about 66% of the standardized residuals have values between -1 and +1, and about 95% of the standardized residuals have values between -2 and +2. A larger standardized residual indicates that the point is far from the regression. Values less than -2.5 or larger than 2.5 may indicate outlying cases. Studentized Residuals. The Studentized residual is a standardized residual that also takes into account the greater confidence of the predicted values of the dependent variable in the middle of the data set. By weighting the values of the residuals of the extreme data points (those with the lowest and highest independent variable values), the Studentized residual is more sensitive than the standardized residual in detecting outliers. This residual is also known as the internally Studentized residual, because the standard error of the estimate is computed using all data. Studentized Deleted Residuals. The Studentized deleted residual, or externally Studentized residual, is a Studentized residual which uses the standard error of the estimate computed after deleting the data point associated with the residual. This reflects the greater effect of outlying points by deleting the data point from the variance computation. The Studentized deleted residual is more sensitive than the Studentized residual in detecting outliers, since the Studentized deleted residual results in much larger values for outliers than the Studentized residual.
Influence Diagnostics
Row. This is the row number of the observation. Cooks Distance. Cooks distance is a measure of how great an effect each point has on the estimates of the parameters in the regression equation. It is a measure of how much the values of the regression coefficients would change if that point is deleted from the analysis. Values above 1 indicate that a point is possibly influential. Cooks distances exceeding 4 indicate that the point has a major effect on the values of the parameter estimates.
Leverage. Leverage values identify potentially influential points. Observations with leverages two times greater than the expected leverages are potentially influential points. The expected leverage of a data point is
where there are p parameters and n data points. Because leverage is calculated using only the dependent variable, high leverage points tend to be at the extremes of the independent variables (large and small values), where small changes in the independent variables can have large effects on the predicted values of the dependent variable. DFFITS. The DFFITSi statistic is a measure of the influence of a data point on regression prediction. It is the number of estimated standard errors the predicted value for a data point changes when the observed value is removed from the data set before computing the regression coefficients. Predicted values that change by more than 2.0 standard errors when the data point is removed are potentially influential.
Figure 14-19 Regression Report Showing the Influence Diagnostics
710 Chapter 14
Row. This is the row number of the observation. Predicted. This column shows the value for the dependent variable predicted by the regression model for each observation. Confidence. The confidence interval for the regression gives the range of variable values computed for the region containing the true relationship between the dependent and independent variables, for the specified level of confidence. The 95% Conf-L values are lower limits and the 95% Conf-U values are the upper limits. Prediction. The confidence interval for the population gives the range of variable values computed for the region containing the population from which the observations were drawn, for the specified level of confidence. The 95% Pred-L values are lower limits and the 95% Pred-U values are the upper limits.
For example, the standard.jfl regression library supplied with SigmaPlot has twelve categories of built-in equations: Polynomial Peak Sigmoidal Exponential Decay Exponential Rise to Maximum Exponential Growth Hyperbola Waveform Power Rational Logarithm 3D Standard Curves
712 Chapter 14
Ligand Binding These categories correspond to the section names within the Standard.jflnotebook. For more information, see Regression Equation Library in Chapter 18. To see the library currently in use, click Back in the Regression Wizard equation panel. Previously selected libraries and open notebooks can be selected from the Library drop-down list.
2. Click Back to view the library panel. To change the library used, enter the new library path and name, or click Browse. The File Open dialog box appears. 3. Change the path and select the file to use as your regression library. When you start the Regression Wizard next, it will continue to use the equation library selected in the library panel. Note: Opening a regression equation directly from a notebook does not reset the equation library.
714 Chapter 14
fitting the data, you can switch the numbers back to dates. For more information, see Setting Day Zero in Chapter 3.
Figure 14-23 You can curve fit dates, but you must convert the dates to numbers first. Time only data (as shown in column 1) does not require a conversion.
Note: If you have entered clock times only, then you can directly curve fit those time without having to convert these to numbers. Time only entries assume the internal start date of 4713 B.C. (the start of the Julian calendar). However, if you have entered times using a more recent calendar date, you must convert these times to numbers as well.
2. Click the Worksheet tab. 3. From the Settings for list, select Date and Time. 4. Set Day Zero to be the first date of your data, or to begin very close to the starting date of your data. You must include the year as well as month and day. 5. Click OK, then view the worksheet and select your data column. 6. From the menus select:
Format Cells
716 Chapter 14
Your dates are converted to numbers. These numbers should be relatively small numbers. If the numbers are large, you did not select a Day Zero near your data starting date. 9. If the axis range of you graph is manual, convert it back to automatic. Select the axis, then open the Graph Properties dialog box and change the range to Automatic. For more information, see Changing Axis Range in Chapter 9. 10. Click you curve and run your regression. When you are finished, you must convert the original and fitted curve x variable columns back to dates.
The Format Cells dialog box appears. 2. Click the Data tab. 3. Under Types, select Date and Time. 4. On the Date drop-down list, click a date format. 5. Click OK.
718 Chapter 14
Figure 14-27 Converting Numeric Data back to Date and Time Data
When the columns are converted back to dates, the graph re-scales and you have completed your date and time curve fit.
Figure 14-28 The Data and Fitted Curve X Variables Converted Back to Dates and Graphed
720 Chapter 14
SigmaPlot uses an estimate of the "condition" of the array (ill conditioned means nearly singular) to generate this message (see Dongarra, J.J., Bunch, J.R., Moler, C.B., and Stewart, G.W., Linpack Users Guide, SIAM, Philadelphia, 1979 for the computation of condition numbers). Usually this message should be taken seriously, as something is usually very wrong. For example, if an exponential underflow has occurred for all x values, part of the equation is essentially eliminated. SigmaPlot still tries to estimate the parameters associated with this phantom part of the equation, which can result in invalid parameter estimates. A minority of the time the "correct," though poorly conditioned, parameters are obtained. This situation may occur, for example, when fitting polynomial or other linear equations. Parameters may not be valid. Array numerically singular on final iteration. This is the limiting case of the above condition where the array cannot be inverted and the condition number is infinite. In this case, the parameter values are not well specified and their standard errors cannont be properly interpreted. Parameters may not be valid. Overflow in partial derivatives. The partial derivatives of the function to be fit, with respect to the parameters, are computed numerically using first order differences. Math errors from various sources can cause errors in this computation. For example, if your model contains exponentials and the parameters and independent variable values cause exponential underflows, then the numerical computation of the partial derivative will be independent of the parameter(s). SigmaPlot checks for this independence. Check the parameter values in the results screen, the range of the independent variable(s) and your equation to determine the problem. There may be inconsistent constraints. Check constraint equations. This occurs if you have defined constraints like a>0 and a<-1.
No observations to fit. The regression cannot proceed unless at least one x,y data pair (observation) is included. Check to be sure that the data columns referenced in the regression specifications contain data. No parameters to fit. The regression specifications do not include any parameter definitions. To add parameter definitions, return to the Equation Options dialog box and type the parameter definitions in the Parameters edit window. No weight statement. The regression specifications include a fit to statement with an unknown weight variable. Check the Variables edit window to see if a weight variable has been defined and that this corresponds to the variable in the regression statement. Not enough or bad number of observations. In regression, the x and y data sets must be of the same size. The data sets (x and y columns) you specified contain unequal numbers of values. Problem loading the file [Filename]. File too long; truncated. The fit file you tried to load is too long. Regression files can be up to 50 characters wide and 80 lines long. Any additional characters or lines were truncated when the file was loaded into the Edit Window. Section has already been submitted. This regression section has already been defined. Symbol [Variable or Function] has not been defined. The fit to statement in the regression definition contains an observed variable which is undefined, or the fit to statement in the regression definition contains an undefined function. Examine the regression specifications you have defined and be sure that the dependent variable listed in the regression statement exists and corresponds to the variable defined in the Variables edit window and that the function listed in the regression statement exists and corresponds to the function you defined in the Equations edit window. Unreferenced variable. The regression specifications define a parameter that is not referenced in any other statements. Either delete the parameter definition, or reference it in another statement.
722 Chapter 14
sum of squares no longer decreases significantly. For more information, see Curvefitting Algorithm above. For complicated curve fitting problems, use SigmaPlots new Dynamic Fit Wizard to find the best solution. The Dynamic Fit Wizard automates the search for initial parameter values that lead to convergence to the best possible solution. For example, typically for a user-defined function, you would need to edit the code to manually enter the initial parameter values, possibly repeatedly, until you find the best fit. For more information, see Entering Initial Parameters in Chapter 15. The Dynamic Fit Wizard takes away the guesswork of estimating the initial parameters. You enter a range for the minimum and maximum initial parameter values, and the Dynamic Fit Wizard does the rest, giving you the confidence that you will find the best fit.
If you are using a worksheet, select the variables in the worksheet you want to fit, then from the menus select:
Statistics Nonlinear Regression Dynamic Fit Wizard
724 Chapter 14
changing initial parameter estimates, parameter constraints, weighting, and other related settings. For more information, see Equation Options above. If you pick variables from a worksheet column, you can also set the data format. For more information, see Variable Options above. 4. When you have selected your variables, you can either click Finish, or click Next to view the Initial Results.
5. Enter or select a number from the Number of fits drop-down list. The default value is 200. This is a good value to start with through for more difficult problems, you may want to increase it. 6. Set the Marquadt-Levenberg algorithm from the Maximum number of iterations dropdown list. Again, more difficult problems may require a larger value. Check the Iterations exceeding percentage in the Dynamic Fit report. If this is greater than 50% then increase the Maximum number of iterations. Once the process exceeds this limit for a fit, then there is "no convergence" for this fit. The process continues with the next set of starting parameter values.
7. Under Select range method select User-defined. 8. Select a parameter listing under Select a parameter to modify its range. 9. Enter the ranges into the Minimum and Maximum boxes. 10. Click Modify. 11. These new values appear under the Minimum and Maximum columns. 12. Click Next. As the dynamic curve fit procedure begins, a progress bar appears in the status bar in the lower left corner of SigmaPlot, indicating the proportion of fits as they are analyzed over time. Once the set Number of Fits has been reached, the Dynamic Fit Wizard - Fit Results panel appears.
726 Chapter 14
1. Select which results you want to keep from the Results for best-fit solution list. These settings are remembered between regression sessions. 2. To select worksheet results options for all converged fits, click More worksheet results. The Result Worksheet Options dialog box appears. For more information, see Worksheet Result Options below. 3. To set the options for the report, click Report. For more information, see Setting Nonlinear Regression Report Options above. 4. Click Next to set the graph options. Worksheet Result Options. You can select several types of calculations to compare and study fit results obtained from the Dynamic Fit Wizard. These results of these calculations appear in the worksheet for every (convergent) solution. Basic results - sum of squares, iteration counts, final parameters. Selected by default. Select this option to create one column for sum of squares, titled "SumSq", one column for the number of iterations needed for convergence, titled "Iterations", and p columns of the final, or best-fit, parameter values, where p is the number of parameters in the fit model. The column title for this is "<parameter value name>-Final". All worksheet results are linked to the values in the sum of squares column. Its values are ordered from smallest to largest. Each row of data refers to a particular fit. Clearing this option disables the Additional options. Additional options. Select to place any of the following options below into the worksheet. To better distinguish columns of data, single blank columns are inserted into the worksheet between the groups of data specified by the options below. Condition number. Select to create a worksheet column called "Condition Number". This is the condition number of the covariance matrix obtained at the final iteration. For a positive-definite matrix, the condition number is computed as the ratio of the maximum and minimum eigenvalues. It is a measure of the sensitivity of the sum of squares value to a change in parameter values at the final iteration. Larger condition numbers indicate more uncertainty in specifying the best-fit parameters. The values in this column use E notation, with three decimal points of precision. For more information, see Engineering and E Notation in Chapter 3.
Starting parameters. These are the initial parameter estimates used to start the fit algorithm. These values are selected from the parameter ranges that you have specified. The title for this column is "<parameter-name>-Start". Parameter standard errors. These are the asymptotic standard errors for the parameters computed at the final iteration. They measure the range of uncertainty in specifying the best-fit parameters. The title of the Parameter standard errors columns is "<parameter name>-StdErr". Coefficients of variation. For each parameter value in the best-fit solution, this is the percent value of the ratio of the parameters standard error to the parameters absolute value. The title for this column is "<parameter name>-CV%". Dependencies. Given a parameter value in the best-fit solution, this is a number between 0 and 1 that measures the dependency of the rate of change of the predicted values with respect to the parameter on the rates of changes with respect to the other parameters. A value close to 1 indicates uncertainty in specifying the given parameter value in the best-fit solution. The title for this column is "<parameter name>-Dep".
728 Chapter 14
6. Select Add equation to graph title to insert the equation of the curve fit under the title of the graph. 7. Select Create a dynamic fit profile graph to create a separate graph that plots the base10 logarithm of the minimum sum of squares of each fit versus the fit index. 8. After selecting the graphed results you want, click Finish. Click Next only if you want to select the specific columns used to contain the data for the fitted curve.
When you click Finish in the Dynamic Fit Wizard, a report appears, divided into the following sections: Dynamic Fit Options. This section displays the dynamic fit options that you chose on the Search Options panel of the Dynamic Fit Wizard. These include the total number of fits, the maximum number of iterations, and the search ranges for each parameter. For more information, see Setting the Dynamic Curve Fit Options above. Summary of Fit Results. This section displays a summary of the outcomes from all of the fits that were executed. As a percent of the total number of fits, the following conditions are reported: Converged. Those fits that satisfied the convergence criterion. Singular Solutions. Those convergent fits whose covariance matrix is singular. Ill-Conditioned Solutions. Those convergent fits whose covariance matrix is illconditioned (to machine precision). Evaluation Failures. Fits that failed to converge due to an evaluation error of the fit equation induced by certain (out of domain) parameter values. Iterations Exceeding <Iterations Exceeding>. Fits that failed to converge after the iteration limit was reached. This user specified limit is inserted into the brackets above. Inner-Loop Failures. Fits where the Levenberg-Marquardt parameter has increased above a prescribed value when searching for a parameter direction to decrease the residual sum of squares. The precision of the percentage value that is displayed for each above condition is determined by the total number of fits specified by the user. This precision (number of places to the right of the decimal point) equals the order of magnitude of the total number of fits minus one. Also, a condition is only listed in the report if there was at least one fit that satisfied it - no zeroes are entered. Results for the Overall Best-Fit Solution. This section is a detailed analysis of the best solution that was found out of the total number of fits. This section mimics the output of an ordinary nonlinear regression report. The fits that were performed either converged or did not converge. Of those that converged, some may be singular and some may be ill-conditioned. Therefore, the combined percentage of singular and ill-conditioned solutions is less than or equal to the percentage of convergent solutions. Also, he combined percentages of convergent solutions, evaluation failures, fits exceeding maximum number of iterations and inner loop failures equaling 100.
730 Chapter 14
curvature parameters since they affect the curvature of the upper and lower portions of the curve, as shown below This equation was implemented in SigmaPlot with preliminary but reasonable initial parameter estimate functions. Experimental blood pressure and heart rate data from the Appendix in the Ricketts and Head paper (1) was fit with this equation. A single curve fit was performed first and this was then followed by a dynamic fit using 500 fits. The dynamic fit found a presumed global minimum and a local minimum. The single fit, using the simple initial parameter estimate functions, found the local minimum but not the presumed global minimum. The dynamic fit profile, shown below, shows the logarithm of the residual sum of squares values, from smallest to largest, for the results of the dynamic fit. The two steps in the insert graph show the values for the presumed global minimum and the first local minimum. The majority of the results (fit numbers between about 30 to 475) corresponded to a straight horizontal line fit at the mean of the y data. The data and the two fit solutions corresponding to the sum of squares values shown in the insert graph are shown below. Notice how the magnitude of Slope1 and Slope2 values switches between the two fit solutions with Slope1 being the smallest magnitude (in absolute value) for the global minimum and the largest magnitude for the local minimum. This is not the artificial switch that occurs for the parameters in the exponents of sums of exponentials (the two exponential terms switching places) since the residual sums of squares values are different for the two solutions. This behavior seems to be a characteristic of this particular 5PL equation as the following example shows. This 5PL equation was investigated further using simulated data. A single fit to the data was performed followed by a dynamic fit using 500 fits. The dynamic fit profile insert graph shows the presumed global minimum and three local minima. The four fit solutions appear in pairs with each pair nearly coinciding with each other. The first pair has nearly a zero slope for small x (dose) values where the second pair does not. The large and small magnitude slopes again switched from Slope1 to Slope 2 in the first pair (global minimum and local minimum 1) and also for the second pair (local minimum 2 and local minimum 3). As previously mentioned the slope of the 5PL curve at x = logEC50 is proportional to the sum of the two slopes. For the first pair of
732 Chapter 14
solutions this sum, Slope 1 + Slope 2, is 2.90 and 2.90 showing how similar are the two curves. This sum for the second pair is also close, 1.82 and 1.68. This example shows: The existence of a local minimum for the 5PL model and actual experimental data. The apparent "pairing" of fit solutions for this 5PL model. The creation of additional local minima with increased data error.
The residual sum of squares decreased as expected for the two-site fit but Akaikes AICc value was so much larger than for the one-site fit that the probability that the onesite fit is correct is an overwhelming 99.97%.
Converged 100.0% Singular Solutions 0.5% All 200 fits converged with only one being singular. It is clear that a dynamic fit was not required for this simple problem. Also see Example 5 for a much more complicated problem for which the fit results summary had very few failures. On the other hand fitting the Bennett5 equation to the data provided at the NIST web site (4) using the "higher level of difficulty" set of starting parameter values results in:
Summary of Fit Results:
734 Chapter 14
Evaluation Failure 0.5% Iterations Exceeding 2000 33.0% Inner-Loop Failures 22.0% This example demonstrates the entire set of dynamic fit result conditions. Only 44.5% of the 200 fits converged. 6% of the total were singular and 31.5% were ill conditioned. Thus 37.5% of the fits were of "low quality"; i.e., either invalid or extremely variable parameter estimates will be obtained in these cases. So only 44.5% - 37.5% = 7% of the 200 fits would enable valid parameter estimates to be obtained. Of the 55.5% that did not converge 0.5% (1 fit) had an evaluation failure (e.g, zero divide, exponential overflow, square root of negative number), 33% exceeded the maximum number of iterations specified (2000 in this case) and 22% had inner loop failures (the number of iterations in the Marquardt-Levenberg algorithm inner loop exceeded 50). This is clearly a difficult curve fitting problem and it shows by using the "higher level of difficulty" initial parameter values how important good initial starting values can be.
Figure 14-29
where f represents ln(viscosity), x1 is temperature and x2 is pressure. For each temperature the data has a roughly linear behavior with pressure (x2). The slope of the pressure data at zero temperature is about 1.5. From the equation for x1 = 0 the slope is approximately 3 + 6 (we assume that the higher order terms in x2 are small corrections since the data is nearly linear). This gives a rough estimate for 3 + 6= 1.5. We dont know how to partition this sum so we will just use as initial estimates 3= 1.0 and 6= 1.0 (rounded up for clarity). A dynamic fit requires a numeric range to be specified for each parameter. We use a rule-of-thumb for the lower and upper limits of the parameter range of the parameter value plus and minus two times the parameter value. So the range for 3and 6 = [-1, 3]. We see from graph above that the data really isnt truly linear with pressure but has some curvature. The 4and 5 terms help account for this. Since this curvature is not very dramatic we will use initial estimates for 4 and 5 that are approximately one tenth and one hundredth of 3, respectively. Since we dont know what the signs of these two parameters should be we let |4| = 0.1 and |5| = 0.01 and let the ranges be symmetric about zero from minus twice to plus twice these values. Therefore the ranges are 4 = [-0.2, 0.2] and 5 = [-0.02, 0.02]. The remaining parameters characterize the behavior with temperature and a coupled behavior with temperature and pressure. If we look at the data in the graph for zero
736 Chapter 14
pressure, x1 = 0, we can get estimates for 1 and 2. The function 1/(2 + x1) is a hyperbolic function that decreases with increasing temperature (x1). When x1 = 2 this function decreases to half its original value. The data for zero pressure decreases from its initial value of over 5.0 to less than 2.5 for some temperature greater than 98.9. So lets use 2 = 100 as a rough initial estimate. The ln(viscosity) is about 5.0 for zero temperature and pressure so we can use 1/2 = 5.0 and get a rough initial estimate 1 = (5.0)*(100) = 500. This gives the ranges for 1 and 2 of [500, 1500] and [-100, 300], respectively. For the coupled term we dont know what the temperature parameter 8 should be so we will just guess that it is roughly the same as the other temperature parameter 2 = 100 in the hyperbolic function. So the range is 8 = [-100, 300]. The remaining two parameters, 7 and 9, are assumed to be small corrections so we will let them be one tenth of the other parameters in their respective sums and, since we dont know their signs, we will let their ranges be symmetric about zero. So 7 will be 0.1*6 or 7 = 0.1 and likewise 9 = 0.1*8 or 9 = 10 and the corresponding ranges are 7 = [-0.2, 0.2] and = [-20, 20]. To summarize, the very rough estimates of initial parameter ranges are: 1 = [-500, 1500] 2 = [-100, 300] 3 = [-1, 3] 4 = [-0.2, 0.2] 5 = [-0.02, 0.02] 6 = [-1, 3] 7 = [-0.2, 0.2] 8 = [-100, 300] 9 = [-20, 20] These ranges were entered into the dynamic fit dialog and 500 fits obtained using a maximum number of iterations equal to 200. The presumed global minimum was found for 14% of the 500 fits. So the technique of using a dynamic fit as a substitute for good initial parameter estimates was successful. The solution found by SigmaPlot was essentially the same as that found by Bates and Watts. SigmaPlot obtained a slightly lower residual sum of squares (0.08738 SigmaPlot vs 0.08996 Bates and Watts). This difference is probably due to the use of different convergence tolerances by the two programs.
The insert graph shows the 69 fit solutions in the first step level that converged to the presumed global minimum. The summary of fit results report is
Summary of Fit Results:
Converged 89.2% Singular Solutions 31.4% Ill-Conditioned Solutions 21.4% Evaluation Failure 0.4% Iterations Exceeding 200 10.0% Inner-Loop Failures 0.4% This shows that nearly 90% of the fits converged and that 36% were well conditioned (89.2 (31.4 + 21.4) = 36.4). Also, there were relatively few failures. Even though this is a complicated equation with nine parameters, the data quality was high enough that good dynamic fit results were obtained. A short transform was written to generate fit curves for each temperature plot in the graph above and this is shown below. The fit is excellent as would be expected from the small residual sum of squares.
738 Chapter 14
Figure 14-30 Monthly averaged atmospheric pressure differences between Easter Island and Darwin, Australia (4). The line is the fit of the sum of sinusoids equation using the SigmaPlot regression wizard.
SigmaPlot found the presumed global minimum, as documented in reference (4), using the one fit from the regression wizard and this is shown in the first graph. Fitting sums of sinusoids to data is known to result in many fit solutions that are local minima. So if you dont have the answer available, how do you know if the solution obtained is the best solution? The answer is that you never do know for sure. Using a dynamic fit, however, can give you additional confidence that you do have the best solution. A dynamic fit of the sums of sinusoids equation
Figure 14-31
was performed using 200 fits and the most difficult starting parameter values listed in reference (4). The dynamic fit profile obtained is shown in the figure below.
740 Chapter 14
Figure 14-32 The dynamic fit profile showing the first four step levels representing the presumed global minimum and three additional local minima.
The multiple fits at each of the first four step levels in the graph above suggest the existence of at least three local minima and the presumed global minimum (at the lowest log sum of squares value). The log sum of squares values for the first four minima are nearly the same, ranging from about 2.90 to 2.98, which suggests that the fit lines corresponding to each solution will be relatively close together. Even though there are nine parameters the dynamic fit report shows this problem to be relatively insensitive to different initial parameter values. All fits converged and relatively few were singular or ill-conditioned.
Summary of Fit Results:
Converged 100.0% Singular Solutions 3.5% Ill-Conditioned Solutions 4.5% The fit solutions corresponding to the four minima are shown in Figure 11.
Figure 14-33 Fit solutions for the presumed global minimum and the three local minima. The corresponding residual sums of squares are shown in the legend.
These solutions are relatively close to one another as the residual sum of squares suggests. The 12 month cycle is obvious in these solutions with two other cycles at about 26 and 44 months also present in the first three fits. The fourth fit solution had an 85 month cycle (about twice the 44 month cycle). This latter solution is one example of cycle doubling (tripling, etc.) as a cause of local minima for sums of sinusoid solutions. At first glance at the results, the investigator should have no faith that the best fit was found. But after performing a dynamic fit where 1) multiple fit solutions were found to occur at the lowest residual sums of squares when starting from different initial parameter values, 2) all fits converged and 3) a collection of relative minima were found that were not the solution found by the curve fitter originally, ones confidence that the fit shown is in fact the best solution is raised considerably. Though specifying a 12-month cycle parameter does not need justification, a dynamic fit with a 10th parameter corresponding to the 12-cycle period was performed and essentially the same presumed global minimum was found. The residual sum of squares was slightly lower at 773 and the 12-month cycle time was estimated to be 11.9 months.
742 Chapter 14
References
(1) Ricketts, J.H. and Head, G.A. A five-parameter logistic equation for investigating asymmetry of curvature in baroreflex studies. Am. J. Physiol.277 (Regulatory Integrative Comp. Physiol. 46); R441-R454, 1999. (2) Aymerich, M.S., Alberdi, E.M., Martinez, A. And Becerra, S.P. Evidence for Pigment Epithelium-Derived Factor Receptors in the Neural Retina. Invest. Ophthal. & Visual Science, Vol 42, No. 13, 3287-3293, 2001. (3) Bates, D.M, and Watts, D.G. Nonlinear Regression Analysis and its Applications. Wiley, New York, 1988, pp 87-89. (4) The NIST Statistical Reference Datasets for Nonlinear Regression, http://www.itl.nist.gov/div898/strd/nls/nls_main.shtml, dataset ENSO.
744 Chapter 14
Note: SigmaPlot remembers the equation for the next time you open the Global Fit Wizard. If the Finish button is available, click it to complete your regression. If it is not available, or if you want to further specify your results, click Next.
746 Chapter 14
3. Select Confidence and prediction bands to display confidence and prediction bands on the graph. For more information, see Confidence and Prediction Bands above. 4. Note: This option only appears if you select either Create new graph or Add to graph. 5. Select Extend fit to axes to extend the equation curve to intersect the Y-axis. 6. Select Add equation to graph title to insert the equation of the curve fit under the title of the graph. 7. After selecting the graphed results you want, click Finish.
Ill-Conditioned Solutions. Those convergent fits whose covariance matrix is illconditioned (to machine precision). Evaluation Failures. Fits that failed to converge due to an evaluation error of the fit equation induced by certain (out of domain) parameter values. Iterations Exceeding <Iterations Exceeding>. Fits that failed to converge after the iteration limit was reached. This user specified limit is inserted into the brackets above. Inner-Loop Failures. Fits where the Levenberg-Marquardt parameter has increased above a prescribed value when searching for a parameter direction to decrease the residual sum of squares. The precision of the percentage value that is displayed for each above condition is determined by the total number of fits specified by the user. This precision (number of places to the right of the decimal point) equals the order of magnitude of the total number of fits minus one. Also, a condition is only listed in the report if there was at least one fit that satisfied it - no zeroes are entered. Results for the Overall Best-Fit Solution. This section is a detailed analysis of the best solution that was found out of the total number of fits. This section mimics the output of an ordinary nonlinear regression report. The fits that were performed either converged or did not converge. Of those that converged, some may be singular and some may be ill-conditioned. Therefore, the combined percentage of singular and ill-conditioned solutions is less than or equal to the percentage of convergent solutions. Also, he combined percentages of convergent solutions, evaluation failures, fits exceeding maximum number of iterations and inner loop failures equaling 100.
748 Chapter 14
Chapter
Editing Code
15
You can edit a regression equation by clicking the Edit Code button in the Regression Wizard or the Dynamic Fit Wizard. This opens the Functions dialog box. Regression equations can be selected from within the wizard, or opened from a notebook directly. You can also create new regression equations. Creating a new equation requires entry of all the code necessary to perform a regression. This chapter covers: Selecting an equation for editing. For more information, see Opening an Existing Equationbelow. Entering equation code. For more information, see Entering Equations below. Defining constants. For more information, see Defining Constants below. Entering variables. For more information, see Entering Variables below. Entering parameters code. For more information, see Initial Parameters below. Entering code for parameter constraints and other options. For more information, see Parameter Constraints below.
749
750 Chapter 15
You can also double-click an equation in a notebook while the Regression Wizard is open to switch to that equation. Once an equation is opened, you can edit it by clicking the Edit Code button.
You can open the Functions dialog box by: Clicking the New button in the Regression Wizard. Choosing File menu New command, and selecting Regression Equation. Right-clicking in the Notebook Manager, and choosing New, Regression Equation from the shortcut menu. A new equation document has no default settings for the equations, parameters, variables, constraints, or other options.
752 Chapter 15
1. On the Statistics menu, click Regression Wizard. 2. Click New to create a new equation. The Function dialog box appears.
To create an equation from the Notebook Manager:
1. Right-click the section where you want the equation to go. If you want the equation to be created in a new section, right-click the notebook icon. 2. On the shortcut menu, click New from the shortcut menu, and then click Equation. The Function dialog box appears.
Copying Equations
You can copy an existing equation from any notebook view to another, and modify it as desired.
This section covers the minimum steps required to enter the code for a regression equation. For more information on entering the code for each section, see: Equations. For more information, see Equations below. Variables. For more information, see Variables below. Weight Variables. For more information, see Weight Variables below. Initial Parameters. For more information, see Initial Parameters below. Constraints. For more information, see Constraints below. Other Options. For more information, see Other Options below.
Adding Comments
Place comments in the edit box by preceding them with an apostrophe (), or a semicolon (;). You can also use apostrophes or semicolons to comment out equations instead of deleting them.
754 Chapter 15
Entering Equations
To enter the code for the Equation section: 1. Click in the Equation window and type the regression equation model, using the transform language operators and functions. The equation should contain all of the variables you plan to use as independent variables, as well as the name for the predicted dependent variable (which is not your y variable). You can use any valid variable name for your equation variables and parameters, but short, single letter names are recommended for the sake of simplicity. Omit the observed dependent variable name from the regression model. The observed dependent variable (typically your y variable) is used in the fit statement. 2. Press the Enter key when finished with the regression equation model, then type the fit statement. The simplest form of the fit statement is: fit f to y Where f is the predicted dependent variable from the regression model, and y is the variable that will be defined as the observed dependent variable (typically the variable plotted as y in the worksheet). You can also define whether or not weighting is used. For more information, see Weight Variables below.
Figure 15-4 Entering the regression equation and the regression statement.
Example: The code f=m*x+b fit f to y can be used as the model for the function, and also defines y as the observed dependent variable. In this example, x is the independent variable, and m and b the equation parameters.
Defining Constants
Constants that appear in the equations can also be defined under the equations heading. If you decide that an equation parameter should be a constant rather than a parameter to be determined by the regression, define the value for that constant here, then make sure you do not enter this value in the parameters section. Constants defined here appear under the Constants option in Equation Options dialog box. For more information, see Equation Options in Chapter 14.
Entering Variables
Independent, dependent, and weighting variables are defined in the Variables section. One of the variables defined must be the observed values of the dependent variable: that is, the "unknown" variable to be solved for. The rest are the independent variables (predictor, or known variables) and an optional weighting variable.
756 Chapter 15
1. Click in the Variables section and type the character or string you used for the first variable in your regression equation. 2. Type an equal sign (=), then enter a range for the variable. Ranges can be any transform language function that produces a range, but typically is simply a worksheet column. Note: The variable values used by the Regression Wizard and Dynamic Fit Wizard depend entirely on what are selected from the graph or worksheet; the values entered here are only used if the From Code data format is selected, or if the regression is run directly from the Function dialog box. Repeat these steps for each variable in your equation. Up to ten independent variables can be defined, but you must define at least one variable for a regression equation to function. The curve fitter checks the variable definitions for errors and for consistency with the regression equation.
Figure 15-5 Entering the variable definitions
Example: To define x and y as the variables for the equation code f=m*x+b, fit f to y, you could enter the code x=col(1), y=col(2), which defines x variable as column 1 and y variable as column 2, using these columns whenever the regression is run directly from the code.
1. Click in the Initial Parameters section and type the name of the first parameter as it appears in your equation model, followed by an equals (=) sign. 2. Enter the initial parameter value used by the curve fitter. Ideally, this should be as close to the real value as possible. This value can be numeric, or a function that computes a good guess for the parameter. Using a function for the initial parameter value is called automatic parameter estimation. For more information, see Automatic Determination of Initial Parameters below. Example: If your data for the equation code f=m*x+b, fit f to y appear to rise to the right and run through the origin, you could define your initial parameter as m=0.5, b=0. These are good initial guesses, since the m coefficient is the slope and the b constant is the y-intercept of a straight line.
758 Chapter 15
Parameter Constraints
Parameter Constraints are completely optional. Use them to limit the parameter ranges to meaningful values for your particular problem. For more information, see Parameter Constraints above.
Options
The Iterations, Step Size and Tolerance options sometimes can be used to improve or limit your curve fit. The default settings work for the large majority of cases, so you do not need to change these setting unless truly required. For more information, see Other Options below.
Saving Equations
Once you are satisfied with the settings you have entered into the Function dialog box, you can save the equation. Clicking OK automatically updates the equation entry in the current notebook or regression library. If you created a new equation, you are prompted to name it before it is added to your notebook. If you are editing an existing equation, you can click Add As to add the code as a new equation to the current library or notebook. In order to save your changes to disk, you must also save the notebook or library. Changes to your current regression library are automatically saved when you close the wizard. You can also save changes before you close the wizard by clicking Save. Click Save As to save the regression library to a new file. If your equation is part of a visible notebook, you can save changes by saving the notebook using the Save button or the File menu Save or Save As commands. Note that when an equation is edited using the Equation Options dialog box, all the changes are also automatically updated and saved.
Equations
The Equation section of the Function dialog box defines the model used to perform the regression as well as the names of the variables and parameters used. The regression equation code is defined using the transform language operators and functions. The equation must contain all of the variables you wish to use. These include all independent variables, the predicted dependent variable, and observed dependent variable. All parameters and constants used are also defined here. The Equation code consists of two required components: The equation model describing the function(s) to be fit to the data. The fit statement, which defines the predicted dependent variable and, optionally, the name of a weighting variable. The independent variable and parameters are defined within the equation function. Also, any constants that are used must also be defined under the Equations section.
Number of Parameters
You can enter and define up to 25 parameters, but a large number of parameters will slow down the regression process. You can determine if you are using too many parameters by examining the parameter dependencies of your regression results. Dependencies near 1.0 (0.999 for example) indicate that the equation is overparameterized, and that you can probably remove one or more dependent parameters. For more information, see Interpreting Fit Results in Chapter 14.
760 Chapter 15
Defining Constants
Define constants by setting one of the parameters of the equation model to a value, using the form constant=value For example, one commonly used constant is pi, defined as pi=3.14159265359
with weight Reciprocal would display the option Reciprocal in the Regressions Options dialog box Fit With Weight list.
Variables
Independent, dependent, and weighting variables are defined in the Variables edit window of the Function dialog box. One of the variables defined must be the observed values of the dependent variable: that is, the unknown variable to be solved for. The rest are the independent variables (predictor, or known variables) and any optional weighting variables. Up to ten independent variables can be defined. To define your variables, select the Variables edit window, then type the variable definitions. You generally need to define at least two variablesone for the dependent variable data, and at least one for the independent variable data.
Variable Definitions
Variable definitions use the form: variable = range You can use any valid variable name, but short, single letter names are recommended for the sake of simplicity (for example, x and y). The range can either be the column number for the data associated with each variable, or a manually entered range. Most typically, the range is data read from a worksheet. The curve fitter uses SigmaPlots transform language, so the notation for a column of data is: col(column,top,bottom) The column argument determines the column number or title. To use a column title for the column argument, enclose the column title in quotation marks. The top and bottom arguments specify the first and last row numbers and can be omitted. The default row numbers are 1 and the end of the column, respectively. If both are omitted, the entire column is used. For example, to define the variable x to be column 1, enter: x = col(1) Data may also be entered directly in the variables section. For example, you can define y and z variables by entering: y = {1,2,4,8,16,32,64}
762 Chapter 15
z = data(1,100) This method can have some advantages. In the above example, the data function was used to automatically generate z values of 1 through 100, which is simpler than typing the numbers into the worksheet. Note: The Regression Wizard generally ignores the default variable settings, although it requires valid variable definitions in order to evaluate an equation. Variables are redefined when the variables are selected from within the wizard. However, you can force the use of the hard-coded variable definitions, either by selecting From Code as the data source, or running the regression directly from the Function dialog box.
User-Defined Functions
Any user-defined functions that are used later in the regression code must be defined in the Variables section.
Concatenating Columns
Constructor notation can be used to concatenate data sets. For example, you may want to fit an equation simultaneously to multiple y columns paired with one x column. If the x data is in column 1 and the y data is in columns 2 through 6, you can enter the following variable statements: x = {col (1), col (1), col (1), col (1), col (1)} y = {col (2), col (3), col (4), col (5), col (6)}
The variable x is then column 1 concatenated with itself four times, and variable y is the concatenation of columns 2 through 6. If the function to be fit is f, then the fit statement fit f to y fits f to the dependent variable values in columns 2 through 6 for the independent variable values in column 1.
Weight Variables
Variables used to perform weighted regressions are known as weight variables. All weight variables must be defined along with other variables in the Variables window.
764 Chapter 15
When weighting is used, the norm that is computed and displayed in the Progress dialog box is:
Figure 15-6
This includes the effect of weighting. The residuals computed are the weighted residuals.
Initial Parameters
The code under the Initial Parameters section of the Function dialog box specifies which equation coefficients and constants to vary and also sets the initial parameter values for the regression. To enter parameters, select the Initial Parameters window, then type the parameters definitions using the form: parameter=initial value All parameters must appear in the equation model. All equation unknowns not defined as variables or constants must be defined in Initial Parameters.
766 Chapter 15
Consider the logistic function as an example. This function has the stretched s shape that changes gradually from a low value to a high value or vice versa. The three parameters for this function determine the high value (a), the x value at which the function is 50% of range of the functions amplitude (x0) and the width of the transition (b). As expressed in the transform language, the function is entered into the Equation window as f=a/(1+exp(-(x-x0)/b)) fit f to y. Noise in the data can lead to significant errors in the estimates of x0 and b. Therefore, a smoothing algorithm is used to reduce the noise in the data and three functions are then used on the smoothed data to obtain the parameter estimates. To estimate the parameter a, the maximum, use the y value. Use the x value at 50% of the amplitude to estimate x0, and the difference between the x values at 75% and 25% of the amplitude is used to estimate b. As entered into the Initial Parameters window, these are: a=max(y) Auto b=xwtr(x,y,.5)/4 Auto x0=x50(x,y,.5) Auto Both the fwhm and xwtr transform functions have been specifically designed to aid the estimation of function parameters. For more information, see Example 1: Curve Fitting Pitfalls in Chapter 16. The Auto comment that follows each parameter is used to identify that parameter value as the Automatic setting from within the Equation Options dialog box. Note that these values may not at all reflect the final values, but they are approximate enough to prevent the curve fitter from finding false or invalid results.
768 Chapter 15
Constraints
Linear parameter constraints are defined under the Constraints section. A maximum of 25 constraints can be entered. Use of constraints is optional. Constraints are used to set limits and conditions for parameter values, restricting the regression search range and improving regression speed and accuracy. Liberal use of constraints in problems which have a relatively large number of parameters is a convenient way to guide the regression and avoid searching in unrealistic regions of parameter space.
Valid Constraints
A constraint must be a linear equation of the parameters using an equality (=) or inequality (< or >). For example, the following constraints for the parameters a, b, c, d, and e are valid: a<1 10*b+c/20 > 2 d-e = 15 a>b+c+d+e whereas a*x<1 is illegal since x is not a constant, and b+c^2>4 d*e=1 are illegal because they are nonlinear. Tip: Although the curve fitter checks the constraints for consistency, you should still examine your constraint definitions before executing the regression. For example, the two constraints: a<1 a>2 are inconsistent. The parameter a cannot be both less than 1 and greater than 2. If you execute a regression with inconsistent constraints, a message appears in the Results dialog box warning you to check your constraint equations.
Other Options
You can use several special options to influence regression operation. The different options can be used to speed up or improve the regression process, but their use is optional. The three options are: Iterations, the maximum number of repeated regression attempts. For more information, see Iterations below. Step Size, the limit of the initial change in parameter values used by the regression as it tries different parameter values. For more information, see Step Size below. Tolerance, one of the conditions that must be met to end the regression process. For more information, see Tolerance below. When the absolute value of the difference between the norm of the residuals from one iteration to the next is less than the tolerance, this condition is satisfied and the regression considered to be complete. Options are entered in the Options section edit boxes. The default values are displayed for new equations. These settings will work for most cases, but can be changed to overcome any problems encountered with the regression, or to perform other tasks, such as evaluating parameter estimation.
Iterations
Setting the number of iterations, or the maximum number of repeated regression attempts, is useful if you do not want to regression to proceed beyond a certain number of iterations, or if the regression exceeds the default number of iterations. The default iteration value is 1.00. To change the number of iterations, simply enter the maximum number of iterations in the Iterations edit box.
770 Chapter 15
Using zero iterations can be very useful for evaluating the effect of changes in parameter values. For example, once you have determined the parameters using the regression, you can enter these values plus or minus a percentage, run the regression with zero iterations, then graph the function results to view the effect of the parameter changes.
Step Size
The initial step size used by the Marquardt-Levenberg algorithm is controlled by the Step Size option. The value of the Step Size option is only indirectly related to changes in the parameters, so only relative changes to the step size value are important. The default step size value is 100. To change the step size value, type a new value into the edit box. The step size number equals the largest step size allowed when changing parameter values. Changing the step size to a much smaller number can be used to prevent the curve fitter from taking large initial steps when searching around suspected minima. For more information, see Example 1: Curve Fitting Pitfalls in Chapter 16. If you are familiar with this algorithm, step size is the inverse of the Marquardt parameter.
Tolerance
The Tolerance option controls the conditions that must be met in order to end the regression process. When the absolute value of the difference between the norm of the residuals from one iteration to the next is less than the tolerance, the regression is considered to be complete. The curve fitter uses two stopping criteria: When the absolute value of the difference between the norm of the residuals (square root of the sum of squares of the residuals), from one iteration to the next, is less than the tolerance value, the iteration stops. When all parameter values stop changing in all significant places, the regression stops. When the tolerance condition has been met, a minimum has usually been found. The default value for tolerance is 1.0e-10. To change the tolerance value, type the required value in the Tolerance edit box. The tolerance number sets the value that must be met to end the iterations.
More precise parameter values can be obtained by decreasing the tolerance value. If there is a sharp sum of squares response surface near the minimum, then decreasing the tolerance from the default value will have little effect. However, if the response surface is shallow about the minimum (indicating a large variability for one or more of the parameters), then decreasing tolerance can result in large changes to parameter values. For more information, see Example 1: Curve Fitting Pitfalls in Chapter 16.
772 Chapter 15
Chapter
16
where x0 is the location of the peak value. 1. Open the Pitfalls worksheet and graph by double-clicking the Pitfalls Graph in the Nonlin.jnb notebook. Note the positions of data points on the curve. 2. Open the Simplified Lorentzian regression equation by double-clicking it in the Regression Examples notebook. The Regression Wizard opens and displays the variables panel. 3. Click one of the symbols on the graph so that the selected variables are Columns 1 and 2. The object is to determine the peak location x0 for the data. Since this data was generated from the Lorentzian function above using x0 = 0, the regression should always find the parameter value x0 = 0.
773
774 Chapter 16
To find x0, the curve fitter computes the sum of squares function:
as a function of the parameter x0. The curve fitter then searches this parameter space for any x0 value where a relative minimum exists. The sum of squares for x0 has two minimaan absolute minimum at x0 = 0 and a relative minimum at x0 = 4.03and a maximum at 2.5. As the curve fitter searches for a minimum, it may stumble upon the local minimum and return an incorrect result. If you start exactly at a maximum, the curve fitter may also remain there.
Figure 16-1 The plot of the sum of squares for the location of the peak value of a Simplified Lorentzian Distribution
1. False convergence caused by a small step size. Click Options. Note that the value of x0 is set to 1000, and the Step Size option is set to 0.000001.
Figure 16-2 The Equation Options dialog box showing step size set to 0.00001
2. Click OK, then click Next. Using the large initial value of x0 and a small step size, the curve fitter takes one small step, finds that there is no change in the sum of squares using the default value for tolerance (0.0001), and declares the tolerance condition is satisfied. The very low slope in the sum of squares at this large x0 value causes the regression to stop.
Figure 16-3 The results using a step size of 0.00001
3. False convergence caused by a large step size and tolerance. Click Back, then click Options. Open the Step Size list and select 100; this is the default step size value.
776 Chapter 16
4. Click OK, then click Next. The curve fitter takes a large step, reaches negative x0 values, and finds a value x0 = -546 for which the tolerance is satisfied.
Figure 16-5 The results using a step size of 100
The sum of squares function asymptotically approaches the same value for both large positive and negative values of x, so the difference of the sum of squares for x0 = 1000 and x0 = -546 is within the default value for the tolerance. 5. Reducing tolerance for a successful convergence. Click Back, then click Options again. Change the Tolerance value to 0.000001, then click OK.
6. Click Next. The regression continues beyond x0 = -546 and successfully finds the absolute minimum at x0 = 0.
Summary
When you used a poor initial parameter value, you needed to use a large initial step size to get the regression started, and you had to decrease the tolerance to keep the regression from stopping prematurely. Poor initial parameters can arise also when using the Automatic method of determining initial parameters as well as when constant values are used. You will now use initial parameter values which result in convergence to a local minimum and a local maximum. 7. Finding a local minimum. Click Back, then click the Options button.
778 Chapter 16
Figure 16-7 The results of using a step size of 100 and tolerance of 0.000001
8. Change the initial value of x0 to 10 using the drop-down Parameter Values list.
Figure 16-8 Changing the initial parameter value to 10 and the tolerance to 0.0001
9. In the Tolerance drop-down list, change the tolerance back to the default value of 0.0001, then click OK. 10. Click Next. The regression converges to x0 = 4.03, which corresponds to the local minimum.
In this example, you know that a local minimum was found by viewing the sum of squares function for the single parameter x0. However, when there are many parameters, it is usually not obvious whether an absolute minimum or a local minimum has been found.
Figure 16-10 Nonlinear regression results
11. Finding a local maximum. Click Back, then click Options. Change the initial parameter value of x0 to 2.5, then click OK. 12. Click Next. Because this initial parameter value happens to correspond to the maximum of the sum of squares function, the regression stops immediately. The slope
780 Chapter 16
is zero within the default tolerance, so the curve fitter falsely determines that a minimum has been found. 13. Finding the absolute minimum. Click Back, then click Options. Change the initial value of x0 to 2.0. 14. Click OK to close the Options dialog box, then click Next to execute the regression. The initial parameter value is reasonably close to the optimum value, so the regression converges to the correct value x0 = 0.0.
Figure 16-11 Nonlinear regression results
These last examples demonstrate how the curve fitter can find a local minimum and even a local maximum using poorly chosen initial parameter values.
The data in the graph displays the compartmental behavior of Xenon in the body. Three behaviors are seen: the wash-in from the blood (rapid rise), the washout from the lung (rapid decrease), and the recirculation of Xenon shunted past the lung (slow decrease).
Figure 16-12 The weighted graph
Least squares curve fitting assumes that the standard deviations of all data points are equal. However, the standard deviation for radioactive decay data increases with the count rate. Radioactive decay data is characterized by a Poisson random process, for which the mean and the variance are equal. Weighting must be used to account for the non-uniform variability in the data. These weights are the reciprocal of the variance of the data. For a Poisson process, the variance equals the mean. You can use the inverse of the measurements as an estimate of the weights. The initial weighting variable only needs to be proportional to the inverse variance. 2. Double-click the Weighted Triple Exponential equation in the Weighted Regression section.
782 Chapter 16
3. Click Edit Code, and examine the Variable value: w = 1/(col)2 This sets w to equal the reciprocal of the data in column 2. Click Cancel to close the dialog box. 4. Click the datapoints to select your variables. To use the w variable as the weighting variable, click Options, and select w as the Fit With Weight value.
Figure 16-14 Selecting a weight variable
5. Click OK to close the dialog box. 6. Click Next to run the regression. The curve fitter finds a solution quickly. 7. Click Finish to complete the regression. 8. What would be the result without weighting? Press F5, then click Next and click Options. Change the weighting to (none), then click OK. 9. Click Finish. The curve fitter goes through many more iterations. When it is completed, view the Weighted graph page. The graph shows the nonlinear regression results with and without weighting. The weighted results fit the very small recirculation data (represented by the third exponential) quite well. However, when weighting is not used, the curve fitter ignored the relatively small values in the recirculation portion of the data, resulting in a poor fit.
Figure 16-15 Comparing the function results of weighted and unweighted nonlinear regression fits
784 Chapter 16
3. View the graph page and select a data point on the graph (x:Column1, y:Column2), then click Next to run the regression. The equation with parameters a1, b1, x0 and c is fit to the data. 4. Click Finish and view the graph page. A smooth transition at the join-point = 5.3 between the line and the one-minus-exponential is shown.
The first fit is made to data over a restricted range, which does not reveal the true nature of the data. 1. Open the Dependencies worksheet and graph by double-clicking the graph page icon in the Dependencies section of the Regression Examples notebook. The data generated from the sum of two exponentials is graphed on a semi-logarithmic scale over
the range 0 to 3.
786 Chapter 16
Figure 16-16 The dependencies graph showing the data for the sum of two exponentials
Although the data is slightly curved, the "break" associated with the two distinct exponentials is not obvious. 2. Right-click the curve on the shortcut menu and click Fit Curve to open the Regression Wizard. 3. Select the Exponential Decay category and the Single, 2 Parameter exponential decay equation, then click Next twice.
Figure 16-17 Selecting the 2 parameter single exponential decay equation
The results show that the dependencies are not near 1.0, indicating that the single exponential parameters, a1 and b1, are not dependent on one another.
4. Click Back twice, and change the equation to the Double, 4 parameter exponential decay equation. Click Next twice.
Figure 16-19 Selecting the 4 parameter double exponential decay equation
The results show that the parameter dependencies for the double exponential are acceptable, indicating that they are unlikely to be dependent, and that using a double exponential produces a better fit (the curve fitter in fact finds the exact parameter values used to generate the data, producing a perfect fit with an R2 of 1).
788 Chapter 16
Figure 16-20 The results of fitting the data to the sum of two exponentials
5. Click Back twice, and change the equation to a Triple, 6 Parameter exponential decay equation. Click Next twice.
Figure 16-21 Selecting the 6 parameter triple exponential decay equation
The results show that the parameter dependencies for a, b, c, and d are 1.00, suggesting that the three exponential model is too complex and that one exponential may be eliminated. 6. Click Cancel when finished.
Figure 16-22 The results of fitting the data to the sum of three exponentials
where x is the dose and f(x ) is the response, then using nonlinear regression, you can find the value for x where:
1. Open the Solving Nonlinear Equations worksheet and graph file by double-clicking the graph page icon in the Solving Nonlinear Equations section of the Nonlin.jnb notebook. Note that the value for x at y = 50 appears to be approximately 150.
790 Chapter 16
Figure 16-23 The Solving Nonlinear Equations Graph, a four parameter logistic curve
3. Examine the regression statements. Note that x is a parameter, y = 0, and the fit equation is modified: f = p1/(1 + exp(p2*(x -p3))) + p4 - 50
Since you are fitting f to y = 0, these statements effectively solve the original problem for x when y = 50. The values for parameters a, b, c, and d were obtained by fitting the four parameter logistic equation to a given set of dose response data. 4. Click Finish to execute the regression. The parameter solution is the unknown x. For this example, x is approximately 149.5.
Figure 16-25 The results of the Solving Nonlinear Equations example
1. Open the Multiple Function worksheet and graph by double-clicking the graph page icon in the Multiple Function section of the Nonlin.jnb notebook. The data points are for three dose responses.
792 Chapter 16
Columns 1 and 2 hold the combined data for the three curves. Column 3 is used to identify the three different data sets. A 0 corresponds to the first dataset, 1 to the second, and 2 to the third.
Figure 16-26 The multiple function graph with three curves
2. Double-click the Multiple Functions Equation. The Regression Wizard opens with the variables panel displayed. Click Edit Code.
Figure 16-27 The multiple function statements
3. Examine the fit statements. The fit equation is an if statement which uses different equations depending on the value of d, which is the data set identifier variable. If d = 0, the data is fit to f1: if d = 1, the data is fit to f2; and if d = 2, the data is fit to f3. The functions share the T and n parameters, but have individual E parameters of E1, E2, and E3. 4. Click Run to execute the regression. The fit proceeds slowly but fits each data set to the separate equation. Click Next to ensure that the Predicted function results are saved to the worksheet, then Next again and make sure no graph is being created. Click Finish to end the fit. 5. To graph the results, you need to create a plot of the predicted results. View the page and select the graph, then create a straight line plot of rows 1-12 of column 1 versus rows 1-12 of the predicted results column.
Figure 16-28 Creating a plot of a restricted data range
6. Create two more line plots of rows 13-23 and 24-34. The results plots appear as three separate curves.
794 Chapter 16
Figure 16-29 A graph of the predicted results of the multiple function equation
When fitted to the data in columns 1 and 2 in the Advanced Techniques worksheet, this equation presents several problems: Parameter identifiability. Very large x values. Very large y error value range. These problems are outlined and solved below. If you want to view the regression functions for this equation, open the Advanced Techniques worksheet and graph in the Advanced Techniques section of the Nonlin.jnb notebook. Double-click the Advanced Techniques Equation to open the Regression Wizard. If you want to run the equation, use the graph of the transformed data.
Overparameterized Equations
The equation has four parameters, a, b, c, and d. The numerator in the exponential: -dx(b+cx) can have identical values for an infinite number of possible parameter combinations. For example, the parameter values: b = c = 1 and d = 2 and the values: b = c = 2 and d = 1 result in identical numerator terms. The curve fitter cannot find a unique set of parameters. The parameters are not uniquely identifiable, as indicated by the large values for variance inflation factor (VIF), and dependency values near 1.0. The solution to this problem is to multiply the d parameter with the other terms to create the equation:
then treat the db and dc terms as single parameters. This reduces the number of parameters to three.
Scaling Large Variable Values
The data used for the fit has enormous x values, around a value of 1 1024 (see column 1 in the worksheet above). These x values appear in the argument of an exponential which is limited to about 700, which is much smaller than 1024. However, when the curve fitter tries to find the parameter values which are multiplied with x, it does not try to keep the argument value within 700. Instead, when the curve fitter varies the parameters, it overflows and underflows the argument range, and does not change the parameter values. The solution to this problem is to scale the x variable and redefine some of the parameters. Multiply and divide each x value by 1 1024 to get:
796 Chapter 16
If you let CD = 1024dc and A = 10-24a, the resulting scaled equation is simplified to:
The exponent argument now does not cause underflows and overflows. The graph of the transformed x data is displayed below the original data.
Figure 16-30
The y values for the data range from very small values to very large values. However, for this problem, we know that the y values do not have the same errorssmaller y values have smaller errors. The curve fitter fits the data by minimizing the sum of the squares of the residuals. Because the squares of the residuals extend over an even larger range than the data, small residual squared numbers are essentially ignored. The solution to this non-uniform error problem is to use weighting, so that all residual squared terms are approximately the same size. Fitting with a weighting variable of 1/y2 (the inverse of y squared), which is proportional to the inverse of the variance of the y data, produces a better fit for low y value data.
798 Chapter 16
Figure 16-31
To see the results of the regression without weighting, open the Options dialog box and change the weighting to (none) before finishing.
Figure 16-32 The graph showing the results of weighted and unweighted nonlinear regressions
Chapter
17
SigmaPlot 11.0 provides many predefined functions, including arithmetic, statistical, trigonometric, and number-generating functions. In addition, you can define functions of your own.
Function Arguments
Function arguments are placed in parentheses following the function name, separated by commas. Arguments must be typed in the sequence shown for each function. You must provide the required arguments for each function first, followed by any optional arguments desired. Any omitted optional arguments are set to the default value. Optional arguments are always omitted from right to left. If only one argument is omitted, it will be the last argument. If two are omitted, the last two arguments are set to the default value. You can use a missing value (i.e., 0/0) as a placeholder to omit an argument.
Example
The col function has three arguments: column, top, and bottom. Therefore, the syntax for the col function is: col(column,top,bottom) The column number argument is required, but the first (top) and last (bottom) rows are optional, defaulting to row 1 as the first row and the last row with data for the last row. col(2) returns the entirety of column 2. col(2,5) returns column 2 from row 5 to the end of the column.
799
800 Chapter 17
col(2,5,100) returns column 2 from row 5 to row 100. col(2,0/0,50) returns column 2 from row 1 to the 50th row in the column.
User-Defined Functions
You can create any user-defined function, consisting of any expression in the transform language, and then refer to it by name. For example, the following transform defines the function dist2pts, which returns the distance between two points dist2pts(x1,y1,x2,y2) = sqrt((x2-x1)^2+(y2-y1)^2). You can then use this custom-defined function, instead of the expression to the right of the equal sign, in subsequent equations. For example, to plot the distances between two sets of XY coordinates, with the first points stored in columns 1 and 2, and the second in columns 3 and 4, enter: col(5) = dist2pts(col(1),col(2),col(3),col(4)) The resulting distances are placed in column 5.
cell. The cell function returns a specific cell from the worksheet. For more information, see cell below. col. The col function returns a worksheet column or portion of a column. For more information, see col below. put into. The put into function places variable or equation results in a worksheet column. For more information, see put into below. subblock. The subblock function returns a specified block of cells from within another block. For more information, see subblock below. Data Manipulation Functions. The data manipulation functions are used to generate non-random data, and to sample, select, and sort data. data. The data function generates serial data. For more information, see data below. if. The if function conditionally selects between two data sets. For more information, see if below. nth. The nth function returns an incremental sampling of data. For more information, see nth below. sort. The sort function rearranges data in ascending order. For more information, see sort below. Trigonometric Functions. SigmaPlot and SigmaStat provide a complete set of trigonometric functions. arccos. This functions returns the arccosine of the specified argument. For more information, see arccos below. arcsin. This functions returns the arcsine of the specified argument. For more information, see arcsin below. arctan. This functions returns the arctangent of the specified argument. For more information, see arctan below. cos. This function returns the cosine of the specified argument. For more information, see cos below. sin. This function returns the sine of the specified argument. For more information, see sin below. tan. This function returns the tangent of the specified argument. For more information, see tan below. cosh. This function returns the hyperbolic cosine of the specified argument. For more information, see cosh below.
802 Chapter 17
sinh. This function returns the hyperbolic sine of the specified argument. For more information, see sinh below. tanh. This function returns the hyperbolic tangent of the specified argument. For more information, see tanh below. Numeric Functions. The numeric functions perform a specific type of calculation on a number or range of numbers and returns the appropriate results. abs. The abs function returns the absolute value. For more information, see abs below. exp. The exp function returns the values for e raised to the specified numbers. For more information, see exp below. factorial. The factorial function returns the factorial for each specified number. For more information, see factorial below. mod. The mod function returns the modulus, or remainder of division, for specified numerators and divisors. For more information, see mod below. ln. The ln function returns the natural logarithm for the specified numbers. For more information, see ln below. log. The log function returns the base 10 logarithm for the specified numbers. For more information, see log below. sqrt. The sqrt function returns the square root for the specified numbers. For more information, see sqrt below. Range Functions. The following functions give information on ranges. count. The count function returns the number of numeric values in a range. For more information, see count below. missing. The missing function returns the number of missing values and text strings in a range. For more information, see missing below. size. The size function returns the number of data points in a range, including all numbers, missing values, and text strings. For more information, see size below. Accumulation Functions. The accumulation functions return values equal to the accumulated operation of the function. diff. The diff function returns the differences of the numbers in a range. For more information, see diff below. sum. The sum function returns the cumulative sum of a range of numbers. For more information, see sum below.
total. The total function returns the value of the total sum of a range. For more information, see total below. Random Generation Functions. The two random number generating functions can be used to create a series of normally or uniformly distributed numbers. gaussian. The Gaussian function is used to generate a series of normally (Gaussian or bell shaped) distributed numbers with a specified mean and standard deviation. For more information, see gaussian below. random. The random function is used to generate a series of uniformly distributed numbers within a specified range. For more information, see random below. Precision Functions. The precision functions are used to convert numbers to whole numbers or to round off numbers. int. The int function converts numbers to integers. For more information, see int below. prec. The prec function rounds numbers off to a specified number of significant digits. round. The round function rounds numbers off to a specified number of decimal places. For more information, see round below. Statistical Functions. The statistical functions perform statistical calculations on a range or ranges of numbers. avg. The avg function calculates the averages of corresponding numbers across ranges. It can be used to calculate the average across rows for worksheet columns. For more information, see avg below. max. The max function returns the largest value in a range. For more information, see max below. min. The min function returns the smallest value in a range. For more information, seemin below. mean. The mean function calculates the mean of a range. For more information, see mean below. runavg. The runavg function produces a range of running averages. For more information, see runavg below. stddev. The stddev function returns the standard deviation of a range. The stderr function calculates the standard error of a range. For more information, see stddev below.
804 Chapter 17
Area and Distance Functions. These functions can be used to calculate the areas and distances specified by X,Y coordinates. Units are based on the units used for X and Y. area. The area function finds the area of a polygon described in X,Y coordinates. For more information, see area below. distance. The distance function calculates the distance of a line whose segments are described in X,Y coordinates. partdist. The partdist function calculates the distances from an initial X,Y coordinate to successive X,Y coordinates in a cumulative fashion. For more information, see partdist below. Curve Fitting Functions. These functions are designed to be used in conjunction with SigmaPlots nonlinear curve fitter, to allow automatic determination of initial equation parameter estimates from the source data. You can use these functions to develop your own parameter determination function by using the functions provided with the Standard Regression Equations library provided with SigmaPlot. ape. This function is used for the polynomials, rational polynomials and other functions which can be expressed as linear functions of the parameters. A linear least squares estimation procedure is used to obtain the parameter estimates. For more information, see ape below. dsinp. This function returns an estimate of the phase in radians of damped sine functions. For more information, see dsinp below. fwhm. This function returns the x width of a peak at half the peaks maximum value for peak shaped functions. For more information, see fwhm below. inv. The inv function generates the inverse matrix of an invertible square matrix provided as a block. For more information, see inv below. lowess. The lowess algorithm is used to smooth noisy data. Lowess means locally weighted regression. Each point along the smooth curve is obtained from a regression of data points close to the curve point with the closest points more heavily weighted. For more information, see lowess below. lowpass. The lowpass function returns smoothed y values from ranges of x and y variables, using an optional user-defined smoothing factor that uses FFT and IFFT. For more information, see lowpass below. sinp. This function returns an estimate of the phase in radians of sinusoidal functions. For more information, see sinp below.
x25. This function returns the x value for the y value 25% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions. For more information, see x25 below. x50. This function returns the x value for the y value 50% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions. For more information, see x50 below. x75. This function returns the x value for the y value 75% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions. For more information, see x75 below. xatymax. This function returns the x value for the maximum y in the range of y coordinates for peak shaped functions. For more information, see xatymax below. xwtr. This function returns x75-x25 for sigmoidal shaped functions. For more information, see xwtr below. Miscellaneous Functions. These functions are specialized functions which perform a variety of operations. choose. The choose function is the mathematical n choose r function. For more information, see choose below. histogram. The histogram function generates a histogram from a range or column of data. For more information, see histogram below. interpolate. The interpolate function performs linear interpolation between X,Y coordinates. For more information, see interpolate below. polynomial. The polynomial function returns results for specified independent variables for a specified polynomial equation. For more information, see polynomial below. rgbcolor. The rgbcolor(r,g,b) color function takes arguments r,g, and b between 0 and 255 and returns color to cells in the worksheet. For more information, see rgbcolor below. Special Constructs. Transform constructs are special structures that allow more complex procedures than functions. for. The for statement is a looping construct used for iterative processing. For more information, see for below. if...then...else. The if...then...else construct proceeds along one of two possible series of procedures based on the results of a specified condition. For more information, see if...then...else below.
806 Chapter 17
Fast Fourier Transform Functions. Use these functions to remove noise from and smooth data using frequency-based filtering. fft. The fft function finds the frequency domain representation of your data. For more information, see fft below. invfft. The invfft function takes the inverse fft of the data produced by the fft to restore the data to its new filtered form. For more information, see invfft below. real. The real function strips the real numbers out of a range of complex numbers. For more information, see real below. img. The img function strips the imaginary numbers out of a range of complex numbers. complex. The complex function converts a block of real and/or imaginary numbers into a range of complex numbers. For more information, see complex below. mulcpx. The mulcpx function multiplies two ranges of complex numbers together. invcpx. The invcpx takes the reciprocal of a range of complex numbers. For more information, see invcpx below. Probability Functions. Use these functions to compute and verify statistical measures such as significant probabilities, critical values of statistics, confidence intervals and histogram comparisons. normdist. This function is the cumulative normal (or Gaussian) distribution function. It returns the probability that a normal random variable is less than a specified independent variable value. For more information, see normdist below. norminv. This function is the inverse cumulative normal (or Gaussian) distribution function. The probability that a normally distributed random variable is less than the return value is equal to the argument you specify. For more information, see norminv below. normden. This function is the normal (or Gaussian) probability density function. The graph of this function is the familiar bell curve. It returns the value of the slope of the cumulative distribution function at the specified argument value. For more information, see normden below. chisquaredist. This function is the cumulative chi-square distribution function. It returns the probability that a chi-square distributed random variable is less than a specified independent variable value. For more information, see chisquaredist below. chisquareinv. This function is the inverse cumulative chi-square distribution function. The probability that a chi-square distributed random variable is less than
the return value is equal to the argument you specify. For more information, see chisquareinv below. chisquareden. This function is the chi-square probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value. For more information, see chisquareden below. tdist. This function is Students T-distribution function. It returns the probability that a T-distributed random variable is less than a specified independent variable value. For more information, see tdist below. tinv. This function is the inverse of Students T-distribution function. The probability that a T-distributed random variable is less than the return value is equal to the argument you specify. For more information, see tinv below. tden. This function is the T-distributions probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value. For more information, see tden below. fdist. This function is the F-distribution function. It returns the probability that an F distributed random variable is less than a specified independent variable value. For more information, see fdist below. finv. This function is the inverse F-distribution function. The probability that an Fdistributed random variable is less than the return value is equal to the argument you specify. For more information, see finv below. fden. This function is the F-distributions probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value. For more information, see fden below.
abs
The abs function returns the absolute value for each number in the specified range.
Syntax
abs(numbers) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
808 Chapter 17
Example
The operation col(2) = abs(col(1)) places the absolute values of the data in column 1 in column 2.
ape
The ape function is used for the polynomials, rational polynomials and other functions which can be expressed as linear functions of the parameters. A linear least squares estimation procedure is used to obtain the parameter estimates. The ape function is used to automatically generate the initial parameter estimates for SigmaPlots nonlinear curve fitter from the equation provided.
Syntax
ape(x range,y range,n,m,s,f) The x range and y range arguments specify the independent and dependent variables, or functions of them (e.g., ln(x)). Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must be the same size. The n argument specifies the order of the numerator of the equation. The m argument specifies the order of the denominator of the equation. n and m must be greater than or equal to 0 (n, m, 0). If m is greater than 0 then n must be less than or equal to m (if m > 0, n m). The s argument specifies whether or not a constant is used. s=0 specifies no constant term y0 in the numerator, s=1 specifies a constant term y0 in the numerator. s must be either 0 or 1. If n = 0, s cannot be 0 (there must be a constant). The number of valid data points must be greater than or equal to n = m = s. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. 0 f 1. If f is omitted, no smoothing is used.
Example
For x = {0,1,2}, y={0,1,4}, the operation col(1)=ape(x,y,1,1,1,0.5] ) places the 3 parameter estimates for the equation
arccos
This function returns the inverse of the corresponding trigonometric function.
Syntax
arccos(numbers) The numbers argument can be a scalar or range. You can also use the abbreviated function name acos. The values for the numbers argument must be within -1 and 1, inclusive. Results are returned in degrees, radians, or grads, depending on the Trigonometric Units selected in the User-Defined Transform dialog box. Any missing value or text string contained within a range is ignored and returned as the string or missing value. The function range (in radians) is arccose 0 to
Example
The operation col(2) = acos(col(1)) places the arccosine of all column 1 data points in column 2.
arcsin
This function returns the inverse of the corresponding trigonometric function.
Syntax
arcsin(numbers)
810 Chapter 17
The numbers argument can be a scalar or range. You can also use the abbreviated function name asin. The values for the numbers argument must be within -1 and 1, inclusive. Results are returned in degrees, radians, or grads, depending on the Trigonometric Units selected in the User-Defined Transform dialog box. Any missing value or text string contained within a range is ignored and returned as the string or missing value. The function range (in radians) is:
Example
The operation col(2) = asin(col(1)) places the arcsine of all column 1 data points in column 2.
arctan
This function returns the inverse of the corresponding trigonometric function.
Syntax
arctan(numbers) The numbers argument can be a scalar or range. You can also use the abbreviated function name atan. The numbers argument can be any value. Results are returned in degrees, radians, or grads, depending on the Trigonometric Units selected in the User-Defined Transform dialog box. The function range (in radians) is:
Example
The operation col(2) = atan(col(1)) places the arctangent of all column 1 data points in column 2. Note: A convenient way of obtaining the value of is = 4 + atan(1).
area
The area function returns the area of a simple polygon. The outline of the polygon is formed by the xy pairs specified in an x range and a y range. The list of points does not need to be closed. If the last xy pair does not equal the first xy pair, the polygon is closed from the last xy pair to the first. The area function only works with simple nonoverlapping polygons. If line segments in the polygon cross, the overlapping portion is considered a negative area, and results are unpredictable.
Syntax
area(x range,y range) The x range argument contains the x coordinates, and the y range argument contains the y coordinates. Corresponding values in these ranges form xy pairs. If the ranges are uneven in size, excess x or y points are ignored.
Example
For the ranges x = {0,1,1,0} and y = {0,0,1,1}, the operation area (x,y) returns a value of 1. The x and y coordinates provided describe a square of 1 unit.
avg
The avg function averages the numbers across corresponding ranges, instead of within ranges. The resulting range is the row-wise average of the range arguments. Unlike the mean function, avg returns a range, not a scalar. The avg function calculates the arithmetic mean, defined as:
The avg function can be used to calculate averages of worksheet data across rows rather than within columns.
812 Chapter 17
Syntax
avg({x1,x2...},{y1,y2...},{z1,z2...}) The x1, y1, and z1 are corresponding numbers within ranges. Any missing value or text string contained within a range returns the string or missing value as the result.
Example
The operation avg({1,2,3},{3,4,5}) returns {2,3,4}. 1 from the first range is averaged with 3 from the second range, 2 is averaged with 4, and 3 is averaged with 5. The result is returned as a range.
block
The block function returns a block of cells from the worksheet, using a range specified by the upper left and lower right cell row and column coordinates.
Syntax
block(column 1,row 1,column 2,row 2) The column 1 and row 1 arguments are the coordinates for the upper left cell of the block; the column 2 and row 2 arguments are the coordinates for the lower right cell of the block. All values within this range are returned. Operations performed on a block always return a block. If column 2 and row 2 are omitted, then the last row and/or column is assumed to be the last row and column of the data in the worksheet. If you are equating a block to another block, then the last row and/or column is assumed to be the last row and column of the equated block (see the following example). All column and row arguments must be scalar (not ranges). To use a column title for the column argument, enclose the column title in quotes; block uses the column in the worksheet whose title matches the string.
Example
The command block(5,1) = -block(1,1,3,24) reverses the sign for the values in the range from cell (1,1) to cell (3,24) and places them in a block beginning in cell (5,1).
blockheight, blockwidth
The blockheight and blockwidth functions return the number of rows or columns, respectively, of a defined block of cells from the worksheet.
Syntax
blockheight(block) blockwidth(block) The block argument can be a variable defined as a block, or a block function statement.
Example
For the statement x = block(2,1,12,10) The operation cell(1,1) = blockheight(x) places the number 10 in column 1, row 1 of the worksheet. The operation cell(1,2) = blockwidth(x) places the number 11 in column 1, row 2 of the worksheet.
cell
The cell function returns the contents of a cell in the worksheet, and can specify a cell destination for transform results.
Syntax
cell (column,row) Both column and row arguments must be scalar (not ranges). To use a column title for the column argument, enclose the column title in quotes; cell uses the column in the worksheet whose title matches the string. Data placed in a cell inserts or overwrites according to the current insert mode.
chisquareden
This function is the chi-square probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value.
814 Chapter 17
Syntax
chisquareden(x,n) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any value for x must be non-negative. The n argument can be any positive integer and equals the degrees of freedom.
Example
The density function can be used to estimate the probability that the values of a chisquare distributed random variable X lie in a small interval. If X has 8 degrees of freedom, then to estimate the probability that the values of X lie between 5 and 5.1, multiply the density of X at 5 by the length of the interval .1: chisquareden(5,8) * .1 = .10688
chisquaredist
This function is the cumulative chi-square distribution function. It returns the probability that a chi-square distributed random variable is less than a specified independent variable value. A chi-square random variable is defined as a sum of squares of independent standard normal distribution variables. The number of normal variables in the sum is called the degrees of freedom. This distribution is used in goodness-of-fit measures. It describes the distribution of sample variance for a set of normally distributed observations and describes the distribution of the residual sum of squares in regression.
Syntax
chisquaredist(x,n)
Example
Suppose a random variable X is chi-square distributed with 11 degrees of freedom. To compute the probability that the values of this variable are less than 5, we calculate:
chisquaredist(5,11) = .06883
chisquareinv
This function is the inverse cumulative chi-square distribution function. The probability that a chi-square distributed random variable is less than the return value is equal to the argument you specify.
Syntax
chisquareinv(x,n) The x argument can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any scalar value for x represents a probability and so must be between 0 and 1. The n argument can be any positive integer and equals the degrees of freedom.
Example
Suppose a chi-square distributed random variable X has 19 degrees of freedom. To compute the median of X, we calculate: chisquareinv(.5,19) = 18.33765
choose
The choose function determines the number of ways of choosing r objects from n distinct objects without regard to order.
Syntax
choose(n,r) For the arguments n and r, r < n and n choose r is defined as:
816 Chapter 17
Example
To create a function for the binomial distribution, enter the equation: binomial(p,n,r) = choose(n,r) * (p^r) * (1-p) ^ (n-r)
col
The col function returns all or a portion of a worksheet column, and can specify a column destination for transform results.
Syntax
col (column,top,bottom) The column argument is the column number or title. To use a column title for the column argument, enclose the title in quotation marks. The top and bottom arguments specify the first and last row numbers, and can be omitted. The default row numbers are 1 and the end of the column, respectively; if both are omitted, the entire column is used. All parameters must be scalar. Data placed in a column inserts or overwrites according to the current insert mode.
complex
Converts a block of real and imaginary numbers into a range of complex numbers.
Syntax
complex (range,range) The first range contains the real values, the second range contains the imaginary values and is optional. If you do not specify the second range, the complex transform returns zeros for the imaginary numbers. If you do specify an imaginary range, it must contain the same number of values as the real value range.
Example
If x = {1,2,3,4,5,6,7,8,9,10}, the operation complex(x) returns {{1,2,3,4,....,9,10}, {0,0,0,0,....,0,0}}. If x = {1.0,-0.75,3.1} and y = {1.2,2.1,-1.1}, the operation complex(x,y) returns {{1.0,-0.75,3.1}, {1.2,2.1,-1.1}}.
cos
This function returns ranges consisting of the cosine of each value in the argument given. This and other trigonometric functions can take values in radians, degrees, or grads. This is determined by the Trigonometric Units selected in the User-Defined Transform dialog box.
Syntax
cos(numbers) The numbers argument can be a scalar or range. If you regularly use values outside of the usual -2p to 2p (or equivalent) range, use the mod function to prevent loss of precision. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
If you choose Degrees as your Trigonometric Units in the User-Defined Transform dialog box, the operation cos({0,60,90,120,180}) returns values of {1,0.5,0,-0.5,-1}.
cosh
This function returns the hyperbolic cosine of the specified argument.
Syntax
818 Chapter 17
Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
The operation x = cosh(col(2)) sets the variable x to be the hyperbolic cosine of all data in column 2.
count
The count function returns the value or range of values equal to the number of nonmissing numeric values in a range. Missing values and text strings are not counted.
Syntax
count(range) The range argument must be a single range (indicated with the {} brackets) or a worksheet column.
data
The data function generates a range of numbers from a starting number to an end number, in specified increments.
Syntax
data(start,stop,step) All arguments must be scalar. The start argument specifies the beginning number and the end argument sets the last number. If the step parameter is omitted, it defaults to 1. The start parameter can be more than or less than the stop parameter. In either case, data steps in the correct direction. Remainders are ignored.
Example
The operation data(10,1,2) returns the values {10,8,6,4,2}. Note: If start and stop are equal, this function produces a number of copies of start equal to step. For example, the operation data(1,1,4) returns {1,1,1,1}.
diff
The diff function returns a range or ranges of numbers which are the differences between a given number in a range and the preceding number. The value of the preceding number is subtracted from the value of the following number. Because there is no preceding number for the first number in a range, the value of the first number in the result is always the same as the first number in the argument range.
Syntax
diff(range) The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Any missing value or text string contained within the range is returned as the string or missing value.
Example
For x = {9,16,7}, the operation diff(x) returns a value of {9,7,-9}. For y = {4,-6,12}, the operation diff(y) returns a value of {4,-10,18}.
dist
The dist function returns a scalar representing the distance along a line. The line is described in segments defined by the X,Y pairs specified in an x range and a y range.
Syntax
820 Chapter 17
The x range argument contains the X coordinates, and the y range argument contains the Y coordinates. Corresponding values in these ranges form X,Y pairs. If the ranges are uneven in size, excess X or Y points are ignored.
Example
For the ranges x ={0,1,1,0,0} and y = {0,0,1,1,0}, the operation dist(x,y) returns 4.0. The X and Y coordinates provided describe a square of 1 unit x by 1 unit y.
dsinp
The dsinp function automatically generates the initial parameter estimates for a damped sinusoidal functions using the FFT method. The four parameter estimates are returned as a vector.
Syntax
dsinp(x range, y range) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must be the same size, and the number of valid data points must be greater than or equal to 3. Note: dsinp is especially used to estimate parameters on waveform functions. This is only useful when this function is used in conjunction with nonlinear regression.
exp
The exp function returns a range of values consisting of the number e raised to each number in the specified range. This is numerically identical to the expression e^(numbers).
Syntax
exp(numbers)
The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
factorial
The factorial function returns the factorial of a specified range.
Syntax
factorial({range}) The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Any missing value or text string contained within a range is ignored and returned as the string or missing value. Non-integers are rounded down to the nearest integer or 1, whichever is larger. For factorial(x): x < 0 returns a missing value, 0 x < 180 returns x!, and x 170 returns +
Example 1
To create a transform equation function for the Poisson distribution, you can type: Poisson(m,x)=(m^x)*exp(-m)/factorial(x)
822 Chapter 17
fden
This function is the F-distributions probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value.
Syntax
fden(x,m,n) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any value for x must be non-negative. The m argument is any positive integer and equals the numerator degrees of freedom. The n argument is any positive integer and equals the denominator degrees of freedom.
Example
The density function can be used to estimate the probability that the values of an Fdistributed random variable F lie in a small interval. If F has a numerator degrees of freedom equal to 3 and a denominator degrees of freedom equal to 14, then to estimate the probability that the values of F lie between 2 and 2.1, multiply the density of F at 2 by the length of the interval .1: fden(2,3,14) * .1 = .014882
fdist
This function is the F-distribution function. It returns the probability that an F distributed random variable is less than a specified independent variable value. An F-distributed random variable is defined as a scaled ratio of a two chi-square variables. The numerator and denominator degrees of freedom of an F-distributed variable equal the degrees of freedom of the corresponding chi-square variables. This distribution is used to test goodness-of-fit in regression problems and for testing the homogeneity of populations for many groups of normally distributed observations.
Syntax
fdist(x,m,n) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any value for x must be non-negative. The m argument is any positive integer and equals the numerator degrees of freedom. The n argument is any positive integer and equals the denominator degrees of freedom.
Example
Suppose an F-distributed random variable F has a numerator degrees of freedom equal to 3 and a denominator degrees of freedom equal to 14. To compute the probability that the values of F exceed 2, we calculate: P( F > 2 ) = 1 P( F < 2 ) = 1 fdist(2,3,14) = .16035
fft
The fft function finds the frequency domain representation of your data using the Fast Fourier Transform.
Syntax
fft(range) The parameter can be a range of real values or a block of complex values. For complex values there are two columns of data. The first column contains the real values and the second column represents the imaginary values. This function works on data sizes of size 2n numbers. If your data set is not 2n in length, the fft function pads 0 at the beginning and end of the data range to make the length 2n. The fft function returns a range of complex numbers.
824 Chapter 17
Example
For x = {1,2,3,4,5,6,7,8,9,10}, the operation fft(x) takes the Fourier transform of the ramp function with real data from 1 to 10 with 3 zeros padded on the front and back and returns a 2 by 16 block of complex numbers.
finv
This function is the inverse F-distribution function. The probability that an Fdistributed random variable is less than the return value is equal to the argument you specify.
Syntax
finv(x,m,n) The x argument can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any scalar value for x represents a probability and so must be between 0 and 1. The m argument is any positive integer and equals the numerator degrees of freedom. The n argument is any positive integer and equals the denominator degrees of freedom.
Example
Suppose an F-distributed random variable F has a numerator degrees of freedom equal to 3 and a denominator degrees of freedom equal to 14. To calculate the tail of this distribution whose probability is .05, we need to find a number f such that P( F > f ) = .05. This is the same as finding f such that P( F < f) = .95. Therefore, we calculate: finv(.95, 3, 14) = 3.34389
for
The for statement is a looping construct used for iterative processing.
Syntax
for loop variable = initial value to end value step increment do equation equation . . . end for Transform equation statements are evaluated iteratively within the for loop. When a for statement is encountered, all functions within the loop are evaluated separately from the rest of the transform. The loop variable can be any previously undeclared variable name. The initial value for the loop is the beginning value to be used in the loop statements. The end value for the loop variable specifies the last value to be processed by the for statement. After the end value is processed, the loop is terminated. In addition, you can specify a loop variable step increment, which is used to skip values when proceeding from the initial value to end value. If no increment is specified, an increment of 1 is assumed. Note: You must separate for, to, step, do, end for, and all condition statement operators, variables and values with spaces. The for loop statement is followed by a series of one or more transform equations which process the loop variable values. Inside for loops, you can: Indent equations. Nest for loops. Note that these conditions are allowed only within for loops. You cannot redefine variable names within for loops.
Example 1
The operation: for i = 1 to size(col(1)) do cell(2,i) = cell(1,i)*i end for multiplies all the values in column 1 by their row number and places them in column 2.
826 Chapter 17
Example 2
The operation: for j = cell(1,1) to cell (1,64) step 2 do col(10) = col(9)^j end for takes the value from cell (1,1) and increments by 2 until the value in cell (1,64) is reached, raises the data in column 9 to that power, and places the results in column 10.
fwhm
The fwhm function returns value of the x width at half-maxima in the ranges of coordinates provided, with optional Lowess smoothing.
Syntax
fwhm(x range, y range,f) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. 0 f 1. If f is omitted, no smoothing is used.
Example
For x = {0,1,2}, y={0,1,4}, the operation col(1)=fwhm(x,y) places the x width at half-maxima 1.00 into column 1.
gaussian
This function generates a specified number of normally (Gaussian or bell shaped) distributed numbers from a seed number, using a supplied mean and standard deviation.
Syntax
gaussian(number,seed,mean,stddev) The number argument specifies how many random numbers to generate. The seed argument is the random number generation seed to be used by the function. If you want to generate a different random number sequence each time the function is used, enter 0/0 for the seed. Enter the same number to generate an identical random number sequence. If the seed argument is omitted, a randomly selected seed is used. The mean and stddev arguments are the mean and standard deviation of the normal distribution curve, respectively. If mean and stddev are omitted, they default to 0 and 1. Note that function arguments are omitted from right to left. If you want to specify a stddev, you must either specify the mean argument or omit it by using 0/0.
Example
The operation gaussian(100) uses a seed of 0 to produce 100 normally distributed random numbers, with a mean of 0.0 and a standard deviation of 1.0.
histogram
The histogram function produces a histogram of the values range in a specified range, using a defined interval set.
Syntax
histogram(range,buckets) The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
828 Chapter 17
The buckets argument is used to specify either the number of evenly incremented histogram intervals, or both the number and ranges of the intervals. This value can be scalar or a range. In both versions, missing values and strings are ignored. If the buckets parameter is a scalar, it must be a positive integer. A scalar buckets argument generates a number of intervals equal to the buckets value. The histogram intervals are evenly sized; the range is the minimum value to the maximum value of the specified range. If the buckets argument is specified as a range, each number in the range becomes the upper bound (inclusive) of an interval. Values from - to the first bucket fall in the first histogram interval, values from > first bucket to second bucket fall in the second interval, etc. The buckets range must be strictly increasing in value. An additional interval is defined to catch any value which does not fall into the defined ranges. The number of values occurring in this extra interval (including 0, or no values outside the range) becomes the last entry of the range produced by histogram function.
Example 1
For col(1) = {1,20,30,35,40,50,60}, the operation col(2) = histogram(col(1),3) places the range {2,3,2} in column 2. The bucket intervals are automatically set to 20, 40, and 60, so that two of the values in column 1 fall under 20, three fall under 40, and two fall under 60.
Example 2
For buckets = {25,50,75}, the operation col(3) = histogram(col(1),buckets)places {2,4,1,0} in col(3). Two of the values in column 1 fall under 25, four fall under 50, one under 75, and no values fall outside the range.
if
The if function either selects one of two values based on a specified condition, or proceeds along a series of calculations bases on a specified condition.
Syntax
The true value and false value arguments can be any scalar or range. For a true condition, the true value is returned; for a false condition, the false value is returned. If the false value argument is omitted, a false condition returns a missing value. If the condition argument is scalar, then the entire true value or false value argument is returned. If the condition argument contains a range, the result is a new range. For each true entry in the condition range, the corresponding entry in the true value argument is returned. For a false entry in the condition range, the corresponding entry in false value is returned. If the false value is omitted and the condition entry is false, the corresponding entry in the true value range is omitted. This can be used to conditionally extract data from a range.
Example 1
The operation col(2) = if(col(1)< 75,FAIL","PASS") reads in the values from column 1, and places the word FAIL in column 2 if the column 1 value is less than 75, and the word PASS if the value is 75 or greater.
Example 2
For the operation y = if(x < 2 or x > 4,99,x), an x value less than 2 or greater than 4 returns a y value of 99, and all other x values return a y value equal to the corresponding x value. If you set x = {1,2,3,4,5}, then y is returned as {99,2,3,4,99}. The condition was true for the first and last x range entries, so 99 was returned. The condition was false for x = 2, 3, and 4, so the x value was returned for the second, third, and fourth x values.
if...then...else
The if...then...else function proceeds along one of two possible series of calculations based on a specified condition.
830 Chapter 17
Syntax
if condition then statement statement... else statement statement... end if
To use the if...then...else construct, follow the if condition then statement by one or more transform equation statements, then specify the else statement(s). When an if...then...else statement is encountered, all functions within the statement are evaluated separately from the rest of the transform. Note: You must separate if, then, and all condition statement operators, variables, and values with spaces. Inside if...then...else constructs, you can: Type more than one equation on a line Indent equations. Nest additional if constructs. Note that these conditions are allowed only within if...else statements. You cannot redefine variable names within an if...then...else construct.
Example
The operations: i = cell(1,1) j = cell(1,2 If i < 1 and j > 1 then x = col(3) else x = col(4) end if sets x equal to column 3 if i is less than 1 and j is greater than 1; otherwise, x is equal to column 4.
imaginary (img)
The imaginary function strips the imaginary values out of a range of complex numbers.
Syntax
If x = {{1,2,3,4,5,6,7,8,9,10}, {0,0,0,....0,0}}, the operation img(x) returns {0,0,0,0,0,0,0,0,0,0}. If x = {{1.0,-0.75, 3.1}, {1.2,2.1,-1.1}}, the operation img(x) returns {1.2,2.1,-1.1}.
int
The int function returns a number or range of numbers equal to the largest integer less than or equal to each corresponding number in the specified range. All numbers are rounded down to the nearest integer.
Syntax
int(numbers) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
832 Chapter 17
interpolate
The interpolate function performs linear interpolation on a set of X,Y pairs defined by an x range and a y range. The function returns a range of interpolated y values from a range of values between the minimum and maximum of the x range.
Syntax
interpolate(x range,y range,range) Values in the x range argument must be strictly increasing or strictly decreasing. The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Missing values and text strings are not allowed in the x range and y range. Text strings in range are replaced by missing values. Extrapolation is not possible; missing value symbols are returned for range argument values less than the lowest x range value or greater than the highest x range value.
Example
For x = {0,1,2}, y = {0,1,4}, and range = data(0,2,.5) (this data operation returns numbers from 0 to 2 at increments of 0.5), the operation col(1) = interpolate(x,y,range) places the range {0.0,0.5,1.0,2.5,4.0} into column 1. If range had included values outside the range for x, missing values would have been returned for those out-of-range values.
inv
The inv function generates the inverse matrix of an invertible square matrix provided as a block.
Syntax
inv(block) The block argument is a block of numbers with real values in the form of a square matrix. The number of rows must equal the number of columns. The function returns
a block of numbers with real values in the form of the inverse of the square matrix provided.
Example
in block (2,7,4,9).
invcpx
This function takes the reciprocal of a range of complex numbers.
Syntax
invcp(block) The input and output are blocks of complex numbers. The invcpx function returns the range 1/c for each complex number in the input block.
Example
If x = complex ({3,0,1}, {0,1,1}), the operation invcpx(x) returns {{0.33333, 0.0, 0.5}, {0.0,-1.0,-0.5}}.
834 Chapter 17
invfft
The inverse fft function (invfft) takes the inverse Fast Fourier Transform (fft) of the data produced by the fft to restore the data to its new filtered form.
Syntax
invfft(block) The parameter is a complex block of spectral numbers with the real values in the first column and the imaginary values in the second column. This data is usually generated from the fft function. The invfft function works on data sizes of size 2n numbers. If your data set is not 2n in length, the invfft function pads 0 at the beginning and end of the data range to make the length 2n. The function returns a complex block of numbers.
Example
ln
The ln function returns a value or range of values consisting of the natural logarithm (base e) of each number in the specified range.
Syntax
ln(numbers) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value. For ln(x): x < 0 returns an error message, and x = 0 returns - The largest value allowed is approximately x < 10309.
Example
log
The log function returns a value or range of values consisting of the base 10 logarithm of each number in the specified range.
Syntax
log(numbers) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value. For log(x): x < 0 returns an error message, x = 0 returns - The largest value allowed is approximately x < 10309.
Example
lookup
The lookup function compares values with a specified table of boundaries and returns either a corresponding index from a one-dimensional table, or a corresponding value from a two-dimensional table.
Syntax
lookup(numbers,x table,y table) The numbers argument is the range of values looked up in the specified x table. The x table argument consists of the upper bounds (inclusive) of the x intervals within the
836 Chapter 17
table and must be ascending in value. The lower bounds are the values of the previous numbers in the table (- for the first interval). You must specify numbers and an x table. If only the numbers and x table arguments are specified, the lookup function returns an index number corresponding to the x table interval; the interval from - to the first boundary corresponds to an index of 1, the second to 2, etc. If a number value is larger than the last entry in x table, lookup will return a missing value as the index. You can avoid missing value results by specifying 1/0 (infinity) as the last value in x table. The optional y table argument is used to assign y values to the x index numbers. The y table argument must be the same size as the x table argument, but the elements do not need to be in any particular order. If y table is specified, lookup returns the y table value corresponding to the x table index value, i.e., the first y table value for an index of 1, the second y table value for an index of 2, etc. Note: The x table and y table ranges correspond to what is normally called a lookup table.
Example 1
For n={-4,11,31} and x={1,10,30}, col(1)=lookup(n,x)places the index values of 1, 3, and -- (missing value) in column 1.
Figure 17-1
-4 falls beneath 1, or the first x boundary; 11 falls beyond 10 but below 30, and 31 lies beyond 30.
Example 2
To generate triplet values for the range {9,6,5}, you can use the expression lookup(data(1/3,3,1/3),data(1,3),{9,6,5}) to return {9,9,9,6,6,6,5,5,5}. This looks up the numbers 1/3, 2/3, 1, 1 1/3, 1 2/3, 2, 2 1/3, 2 2/3, and 3 using x table boundaries 1, 2, and 3 and corresponding y table values 9, 6, and 5.
Figure 17-2
lowess
The lowess function returns smoothed y values as a range from the ranges of x and y variables provided, using a user-defined smoothing factor. "Lowess" means locally weighted regression. Each point along the smooth curve is obtained from a regression of data points close to the curve point with the closest points more heavily weighted.
Syntax
838 Chapter 17
The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must be the same size, and the number of valid data points must be greater than or equal to 3. The f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. 0 f 1. Note that unlike lowpass, lowess requires an f argument.
Example
For x = {1,2,3,4}, y={0.13, 0.17, 0.50, 0.60}, the operation col(1)=lowess(x,y,1) places the smoothed y data 0.10, 0.25, 0.43, 0.63 into column 1.
lowpass
The lowpass function returns smoothed y values from ranges of x and y variables, using an optional user-defined smoothing factor that uses FFT and IFFT.
Syntax
lowpass(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must be the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines whether FFT and IFFT are used. f must be greater than or equal to 0 and less than or equal to 100 . If f is omitted, no Fourier transformation is used. Note: lowpass is especially designed to perform smoothing on waveform functions as a part of nonlinear regression.
Example
For x = {0,1,2}, y={0,1,4}, the operation col(1)=lowpass(x,y,88) places the newly smoothed data 0.25, 1.50, 2.25 into column 1.
max
The max function returns the largest number found in the range specified.
Syntax
max(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
Example
For x = {7,4,-4,5}, the operation max(x) returns a value of 7, and the operation min(x) returns a value of -4.
mean
The mean function returns the average of the range specified. Use this function to calculate column averages (as opposed to using the avg function to calculate row averages). The mean function calculates the arithmetic mean, defined as:
840 Chapter 17
Syntax
mean(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
Example
min
The min function returns the smallest number in the range specified.
Syntax
min(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
Example
For x = {7,4,-4,5}, the operation max(x) returns a value of 7, and the operation min(x)returns a value of -4.
missing
The missing function returns a value or range of values equal to the number of missing values and text strings in the specified range.
Syntax
missing(range)
The range argument must be a single range (indicated with the { } brackets) or a worksheet column.
mod
The mod function returns the modulus (the remainder from division) for corresponding numbers in numerator and divisor arguments. This is the real (not integral) modulus, so both ranges may be nonintegral values.
Syntax
mod(numerator,divisor) The numerator and divisor arguments can be scalars or ranges. Any missing value or text string contained within a range is returned as the string or missing value. For any divisor 0, the mod function returns the remainder of
For mod(x,0), that is, for divisor = 0, x > 0 returns + x = 0 returns + x < 0 returns -
Example
The operation mod({4,5,4,5},{2,2,3,3}) returns the range {0,1,1,2}. These are the remainders for 4 2, 5 2, 4 3, and 5 3.
mulcpx
The mulcpx function multiplies two blocks of complex numbers together.
Syntax
mulcpx(block, block)
842 Chapter 17
Both input blocks should be the same length. The mulcpx function returns a block that contains the complex multiplication of the two ranges.
Example
normden
This function is the normal (or Gaussian) probability density function. The graph of this function is the familiar "bell curve". It returns the value of the slope of the cumulative distribution function at the specified argument value.
Syntax
normden(x,m,s) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. The m argument can be any number and equals the mean of the distribution. The s argument can be any positive number and equals the standard deviation of the distribution.
Example
The density function can be used to estimate the probability that the values of a normally distributed random variable X lie in a small interval. If X has mean 0 and standard deviation 1, then to estimate the probability that the values of X lie between .5 and .6, multiply the density of X at .5 by the length of the interval .1: normden(.5,0,1) * .1 = .03521
normdist
This function is the cumulative normal (or Gaussian) distribution function. It returns the probability that a normal random variable is less than a specified independent variable value.
Syntax
normdist(x,m,s) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. The m argument can be any number and equals the mean of the distribution. The s argument can be any positive number and equals the standard deviation of the distribution. A normal distribution is called standard if the mean is 0 and the standard deviation is 1.
Example
Suppose a random variable X is normally distributed withmean .5 and standard deviation 2. Then to compute the probability that its values lie between -1 and 1, we calculate: P(-1 < X < 1) = P(X < 1) P(X < -1) 1,.5,2) = .59871 - .22663 = .37208 = normdist(1,.5,2) normdist(-
norminv
This function is the inverse cumulative normal (or Gaussian) distribution function. The probability that a normally distributed random variable is less than the return value is equal to the argument you specify.
Syntax
norminv(x,m,s) The x argument can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any scalar value for x represents a probability and so must be between 0 and 1. The m argument can be any number and equals the mean of the distribution. The s argument can be any positive number and equals the standard deviation of the distribution.
844 Chapter 17
Example
Suppose a random variable X is normally distributed with mean .5 and standard deviation 2. To compute the .25 quartile of X, we calculate: norminv(.25,.5,2) = -.84898
nth
The nth function returns a sampling of a provided range, with the frequency indicated by a scalar number. The result always begins with the first entry in the specified range.
Syntax
nth(range,increment) The range argument is either a specified range (indicated with the {} brackets) or a worksheet column. The increment argument must be a positive integer.
Example
The operation col(1)=nth({1,2,3,4,5,6,7,8,9,10},3) places the range {1,4,7,10} in column 1. Every third value of the range is returned, beginning with 1.
partdist
The partdist function returns a range representing the distance from the first X,Y pair to each other successive pair. The line segment X,Y pairs are specified by an x range and a y range. The last value in this range is numerically the same as that returned by dist, assuming the same x and y ranges.
Syntax
partdist(x range,y range) The x range argument specifies the x coordinates, and the y range argument specifies the y coordinates. Corresponding values in these ranges form xy pairs. If the ranges are uneven in size, excess x or y points are ignored.
Example
For the ranges x = {0,1,1,0,0} and y = {0,0,1,1,0}, the operation partdist(x,y) returns a range of {0,1,2,3,4}. The X and Y coordinates provided describe a square of 1 unit x by 1 unit y.
polynomial
The polynomial function returns the results for independent variable values in polynomials. Given the coefficients, this function produces a range of y values for the corresponding x values in range. The function takes one of two forms. The first form has two arguments, both of which are ranges. Values in the first range are the independent variable values. The second range represents the coefficients of the polynomial, with the constant coefficient listed first, and the highest order coefficient listed last. The second form accepts two or more arguments. The first argument is a range consisting of the independent variable values. All successive arguments are scalar and represent the coefficients of a polynomial, with the constant coefficient listed first and the highest order coefficient listed last.
Syntax
polynomial(range,coefficents) or polynomial(range,a0,a1,...,an)
The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Text strings contained within a range are returned as a missing value. The coefficients argument is a range consisting of the polynomial coefficient values, from lowest to highest. Alternately, the coefficients can be listed individually as scalars.
Example
To evaluate the polynomial y = x2 + x + 1 for x values of 0, 1, and 2, type the equation polynomial({0,1,2},1,1,1). Alternately, you could set x ={1,1,1}, then enter polynomial({0,1,2},x). Both operations return a range of {1,3,7}.
846 Chapter 17
prec
The prec function rounds a number or range of numbers to the specified number of significant digits, or places of significance. Values are rounded to the nearest integer; values of exactly 0.5 are rounded up.
Syntax
prec(numbers,digits) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value. If the digits argument is a scalar, all numbers in the range have the same number of places of significance. If the digits argument is a range, the number of places of significance vary according to the corresponding range values. If the size of the digits range is smaller than the numbers range, the function returns missing values for all numbers with no corresponding digits.
Example
For x = {13570,3.141,.0155,999,1.92}, the operation prec(x,2) returns {14000,3.100,.0160,1000,1.90}. For y = {123.5,123.5,123.5,123.5}, the operation prec(y,{1,2,3,4}) returns {100.0, 120.0,124.0,123.5}.
put into
The put into function places calculation results in a designated column on the worksheet. It operates faster than the equivalent equality relationship.
Syntax
The results argument can be either the result of an equation, function or variable. The column argument is either the column number of the destination column, or the column title, enclosed in quotes. Data put into columns inserts or overwrites according to the current insert mode.
Example
To place the results of the equation y = data(1,100) in column 1, you can type col(1) = y. However, entering put y into col(1) runs faster.
random
This function generates a specified number of uniformly distributed numbers within the range. Rand and rnd are synonyms for the random function.
Syntax
random(number,seed,low,high) The number argument specifies how many random numbers to generate. The seed argument is the random number generation seed to be used by the function. If you want to generate a different random number sequence each time the function is used, enter 0/0 for the seed. If the seed argument is omitted, a randomly selected seed is used. The low and high arguments specify the beginning and end of the random number distribution range. The low boundary is included in the range. If low and high are omitted, they default to 0 and 1, respectively. Note: Function arguments are omitted from right to left. If you want to specify a high boundary, you must specify the low boundary argument first.
Example
The operation random(50,0/0,1,7) produces 50 uniformly distributed random numbers between 1 and 7. The sequence is different each time this random function is used.
848 Chapter 17
real
The real function strips the real values from a complex block of numbers.
Syntax
If x = complex ({1,2,3,...,9,10}, {0,0,...,0}), the operation real(x)returns {1,2,3,4,5,6,7,8,9,10}, leaving the imaginary values out.
rgbcolor
The transform function rgbcolor takes arguments r, g, and b between 0 and 255 and returns the corresponding color to cells in the worksheet. This function can be used to apply custom colors to any element of a graph or plot that can use colors chosen from a worksheet column.
Syntax
rgbcolor(r,g,b) The r,g,b arguments define the red, green, and blue intensity portions of the color. These values must be scalars between 0 and 255. Numbers for the arguments less than 0 or greater than 255 are truncated to these values.
Example
The operation rgbcolor(255,0,0) returns red. The operation rgbcolor(0,255,0) returns green. The operation rgbcolor(0,0,255) returns blue. The following statements place the secondary colors yellow, magenta, and cyan into rows 1, 2, and 3 into column 1:
cell(1,1)=rgbcolor(255,255,0) cell(1,2)=rgbcolor(255,0,255) cell(1,3)=rgbcolor(0,255,255) Shades of gray are generated using equal arguments. To place black, gray, and white in the first three rows of column 1: cell(1,1)=rgbcolor(0,0,0) cell(1,3)=rgbcolor(255,255,255)cell(1,2)=rgbcolor(127,127,127)
round
The round function rounds a number or range of numbers to the specified decimal places of accuracy. Values are rounded up or down to the nearest integer; values of exactly 0.5 are rounded up.
Syntax
round(numbers,places) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value. If the places argument is negative, rounding occurs to the left of the decimal point. To round to the nearest whole number, use a places argument of 0.
Example
The operation round(92.1541,2) returns a value of 92.15. The operation round(0.19112,1) returns a value of 0.2. The operation round(92.1541,-2) returns a value of 100.0.
runavg
The runavg function produces a range of running averages, using a window of a specified size as the size of the range to be averaged. The resulting range is the same length as the argument range.
850 Chapter 17
Syntax
runavg(range,window) The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Any missing value or text string contained within a range is replaced with 0. If the window argument is even, the next highest odd number is used. The tails of the running average are computed by appending
The operation runavg({1,2,3,4,5},3) returns {1.33,2,3,4,4.67}. The value of the window argument is 3, so the first result value is calculated as:
sin
This function returns ranges consisting of the sine of each value in the argument given. This and other trigonometric functions can take values in radians, degrees, or grads. This is determined by the Trigonometric Units selected in the User-Defined Transform dialog box.
Syntax
sin(numbers)
The numbers argument can be a scalar or range. If you regularly use values outside of the usual -2 to 2 (or equivalent) range, use the mod function to prevent loss of precision. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
If you choose Degrees as your Trigonometric Units in the Transform dialog box, the operation sin({0,30,90,180,270}) returns values of {0,0.5,1,0,-1}.
sinh
This function returns the hyperbolic sine of the specified argument.
Syntax
sinh(numbers) The numbers argument can be a scalar or range. Like the circular trig functions, this function also accepts numbers in degrees, radians, or grads, depending on the units selected in the User-Defined Transform dialog box.
Example
The operation x = sinh(col(3)) sets the variable x to be the hyperbolic sine of all data in column 3.
sinp
The sinp function automatically generates the initial parameter estimates for a sinusoidal functions using the FFT method. The three parameter estimates are returned as a vector.
852 Chapter 17
Syntax
sinp(x range, y range) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must be the same size, and the number of valid data points must be greater than or equal to 3. Tip: sinp is especially used to perform smoothing on waveform functions, used in determination of initial parameter estimates for nonlinear regression.
size
The size function returns a value equal to the total number of elements in the specified range, including all numbers, missing values, and text strings. Note that size (X) 1/2 count (X) + missing (X).
Syntax
size(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column.
sort
This function can be used to sort a range of numbers in ascending order, or a range of numbers in ascending order together with a block of data.
Syntax
sort(block,range) The range argument can be either a specified range (indicated with the { } brackets) or a worksheet column. If the block argument is omitted, the data in range is sorted in ascending order.
Example 1
The operation col(2) = sort(col(1)) returns the contents of column 1 arranged in ascending order and places it in column 2. To reverse the order of the sort, you can create a custom function: reverse(x) = x[data(size(x),1)] then apply it to the results of the sort. For example, reverse(sort(x)) sorts range x in descending order.
Example 2
The operation: block(3,1) = sort(block(1,1,2,size(col(2))),col(2)) sorts data in columns 1 and 2 using column 2 as the key column and places the sorted data in columns 3 and 4.
sqrt
The sqrt function returns a value or range of values consisting of the square root of each value in the specified range. Numerically, this is the same as {numbers}^0.5, but uses a faster algorithm.
Syntax
sqrt(numbers) The numbers argument can be a scalar or range of numbers. Any missing value or text string contained within a range is ignored and returned as the string or missing value. For numbers < 0, sqrt generates a missing value.
Example
854 Chapter 17
stddev
The stddev function returns the standard deviation of the specified range, as defined by:
Syntax
stddev(range) The range argument must be a single range (indicated with the {} brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
Example
For the range x = {1,2}, the operation stddev(x) returns a value of .70711.
stderr
The stderr function returns the standard error of the mean of the specified range, as defined by
stderr(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Any missing value or text string contained within a range is ignored.
Example
For the range x = {1,2}, the operation stderr(x) returns a value of 0.5.
subblock
The subblock function returns a block of cells from within another previously defined block of cells from the worksheet. The subblock is defined using the upper left and lower right cells of the subblock, relative to the range defined by the source block.
Syntax
subblock (block, column 1, row 1, column 2, row 2) The block argument can be a variable defined as a block, or a block function statement. The column 1 and row 1 arguments are the relative coordinates for the upper left cell of the subblock with respect to the source block. The column 2 and row 2 arguments are the relative coordinates for the lower right cell of the subblock. All values within this range are returned. Operations performed on a block always return a block. If column 2 and row 2 are omitted, then the last row and/or column is assumed to be the last row and column of the source block. All column and row arguments must be scalar (not ranges).
Example
For x = block (3,1,20,42) the operation subblock (x,1,1,1,1) returns cell (3,1) and the operation subblock (x,5,5) returns the block from cell (7, 5) to cell (20, 42).
sum
The function sum returns a range of numbers representing the accumulated sums along the list. The value of the number is added to the value of the preceding cumulative sum. Because there is no preceding number for the first number in a range, the value of the first number in the result is always the same as the first number in the argument range.
856 Chapter 17
Syntax
sum(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Any text string or missing value contained within the range is returned as the string or missing value.
Example
For x = {2,6,7}, the operation sum(x) returns a value of {2,8,15}. For y = {4,12,-6}, the operation sum(y) returns a value of {4,16,10}.
tan
This function returns ranges consisting of the tangent of each value in the argument given. This and other trigonometric functions can take values in radians, degrees, or grads. This is determined by the Trigonometric Units selected in the User-Defined Transform dialog box.
Syntax
tan(numbers) The numbers argument can be a scalar or range. If you regularly use values outside of the usual -2 to 2 (or equivalent) range, use the mod function to prevent loss of precision. Any missing value or text string contained within a range is ignored and returned as the string or missing value.
Example
If you choose Degrees as your Trigonometric Units in the transform dialog box, the operation tan({0,45,135,180}) returns values of {0,1,-1,0}.
tanh
This function returns the hyperbolic tangent of the specified argument.
Syntax
The operation x = tanh(col(3)) sets the variable x to be the hyperbolic tangent of all data in column 3.
tden
This function is the T-distributions probability density function. It returns the value of the slope of the cumulative distribution function at the specified argument value.
Syntax
tden(x,n) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. The n argument can be any positive integer and equals the degrees of freedom.
Example
The density function can be used to estimate the probability that the values of a Tdistributed random variable T lie in a small interval. If T has 16 degrees of freedom, then to estimate the probability that the values of T lie between 1 and 1.1, multiply the density of T at 1 by the length of the interval .1: tden(1,16) * .1 = .02346
tdist
This function is Students T-distribution function. It returns the probability that a Tdistributed random variable is less than a specified independent variable value.
858 Chapter 17
A T-distributed random variable is defined as a scaled ratio of a standard normal variable and a chi-square variable. The degrees of freedom of a T-distributed variable is defined to be the degrees of freedom of the chi-square variable in the denominator. This distribution is used in computing confidence intervals and for testing the homogeneity of populations for two groups of normally distributed observations.
Syntax
tdist(x,n) The x argument represents the independent variable and can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. The n argument can be any positive integer and equals the degrees of freedom.
Example
Suppose T is a T-distributed random variable with 14 degrees of freedom. To compute the probability that the absolute values of T exceed 2, we calculate: P( |T| > 2 ) = P( T > 2) + P( T < 2) = 2*P( T > 2) = 2*( 1 P( T < 2) ) tdist(2,14)) = .06529 = 2*(1-
This is a typical calculation that is used to test whether two normally distributed groups of observations have the same mean. In this context, the value 2 in our example is called the critical value and is equal to the absolute difference in the sample means of the two groups divided by the pooled standard deviation of the groups. The resulting probability, .06529, is called the probability of significance.
tinv
This function is the inverse of Students T-distribution function. The probability that a T-distributed random variable is less than the return value is equal to the argument you specify.
Syntax
tinv(x,n) The x argument can either be a scalar or a range of numbers. If x is a range, then it must be defined by either using braces { } or by specifying a worksheet column. Any scalar value for x represents a probability and so must be between 0 and 1. The n argument can be any positive integer and equals the degrees of freedom.
Example
Suppose a T-distributed random variable T has 23 degrees of freedom. .75 quartile of T, we calculate: tinv(.75,23) = .68531
total
The function total returns a single value equal to the total sum of all numbers in a specified range. Numerically, this is the same as the last number returned by the sum function.
Syntax
total(range) The range argument must be a single range (indicated with the { } brackets) or a worksheet column. Missing values and text strings contained within the range are ignored.
Example
For x = {9,16,7}, the operation total(x) returns a value of 32. For y = {4,12,-6}, the operation total(y) returns a value of 10.
x25
The x25 function returns an interpolated value of the x data at:
860 Chapter 17
in the ranges of coordinates provided, with optional Lowess smoothing. This is typically used to return the x value for the y value at 25% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions.
Syntax
x25(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. (0 f 1). If f is omitted, no smoothing is used.
Example
x50
The x50 function returns an interpolated value of the x data at:
in the ranges of coordinates provided, with optional Lowess smoothing. This is typically used to return the x value for the y value at 50% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions.
Syntax
x50(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. (0 f 1). If f is omitted, no smoothing is used.
Example
x75
The x75 function returns an interpolated value of the x data at:
in the ranges of coordinates provided, with optional Lowess smoothing. This is typically used to return the x value for the y value at 75% of the distance from the minimum to the maximum of smoothed data for sigmoidal shaped functions.
862 Chapter 17
Syntax
x75(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. (0 f 1). If f is omitted, no smoothing is used.
Example
xatymax
The xatymax function returns the interpolated x value at the maximum y value found, with optional Lowess smoothing.
Syntax
xatymax(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point. x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the
fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. 0 f 1. If f is not defined, no smoothing is used. Note: If duplicate y maximums are found xatymax will return the average value of all the x at y maximums.
Example
For x = {0,1,2}, y={0,1,4}, the operation col(1)=xatymax(x,y) places the x at the y maximum as 2.00 into column 1.
xwtr
The xwtr function returns value of x75-x25 in the ranges of coordinates provided, with optional Lowess smoothing.
Syntax
xwtr(x range, y range, f ) The x range argument specifies the x variable, and the y range argument specifies the y variable. Any missing value or text string contained within one of the ranges is ignored and will not be treated as a data point.x range and y range must have the same size, and the number of valid data points must be greater than or equal to 3. The optional f argument defines the amount of Lowess smoothing, and corresponds to the fraction of data points used for each regression. f must be greater than or equal to 0 and less than or equal to 1. 0 f 1. If f is omitted, no smoothing is used.
Example
For x = {0,1,2}, y={0,1,4}, the operation col(1)=xwtr(x,y) places the x75-x25 as double 1.00 into column 1.
864 Chapter 17
Chapter
18
This appendix lists the equations found in the Regression Equation Library.
Polynomial
Linear
Figure 18-1
Quadratic
Figure 18-2
865
866 Chapter 18
Cubic
Figure 18-3
Peak
Four Parameter Gaussian
Figure 18-7
868 Chapter 18
Sigmoidal
Three Parameter Sigmoid
Figure 18-19
870 Chapter 18
872 Chapter 18
Exponential Decay
Two Parameter Single Exponential Decay
Figure 18-32
874 Chapter 18
Exponential Growth
One Parameter Single Exponential Growth
Figure 18-46
876 Chapter 18
Stirling Model
Figure 18-53
878 Chapter 18
Hyperbola Library
Two Parameter Rectangular Hyperbola
Figure 18-57
Modified Hyperbola I
Figure 18-64
Modified Hyperbola II
Figure 18-65
880 Chapter 18
Waveform
Three Parameter Sine
Figure 18-67
Modified Sin
Figure 18-73
882 Chapter 18
Power
Two Parameter
Figure 18-76
Three Parameter
Figure 18-77
Pareto Function
Figure 18-78
Modified Pareto
Figure 18-83
884 Chapter 18
Rational
One Parameter Rational I
Figure 18-84
886 Chapter 18
Logarithm
Two Parameter I
Figure 18-100
Two Parameter II
Figure 18-101
888 Chapter 18
Second Order
Figure 18-103
Third Order
Figure 18-104
3 Dimensional
Plane
Figure 18-105
Paraboloid
Figure 18-106
Gaussian
Figure 18-107
Lorentzian
Figure 18-108
Standard Curves
Linear Curve
Figure 18-109
890 Chapter 18
Ligand Binding
One Site Saturation
Figure 18-111
892 Chapter 18
Piecewise
Two Segment Linear
Figure 18-123
894 Chapter 18
2 Chapter
2 Chapter
Index
.TIFF files post-processing 510 .TXT files opening 38, 39 .WK* files opening 38, 39
Symbols
.ASC files opening 38, 39 .CVS files opening 38, 39 .DBF files opening 38, 39 .DIF files opening 38, 39 .FIT files 670 Adding to library or notebook 670 .ini files 7 .jfl files 7 .jgg files 7 .JNB files exporting as non-notebook files 237 .jnb files saving 33 .JNT files defined 2 .jnt files 7 .MOC files opening 38, 39 .PRN files opening 38, 39 .SMB files importing 51 .SP5 files opening 38, 39 .SPG files opening 38, 39 .SPW files opening 38, 39
Numerics
10th/90th percentiles box plots 341 1st derivative computing 537 2D graphs adding plots 170 area plots 310, 343, 353 arranging data 139 asymmetric error bar plots 139 asymmetric error bars 322 bar charts 310 box plots 311, 341 column averaged error bar plots 139 creating 311 creating multiple axes for single plot 372 creating plots with asymmetric error bars 322 creating plots with error bars 317 error bars 317 examples 15, 21 grouped bar charts 333, 334 grouped bar charts with error bars 333 linear regression lines 220, 223 modifying plots 220, 227 multiple axes 372 multiple plots 170 plotting data 139, 310, 311, 339 plotting multiple curves 138
611
612 Index
plotting mutiple curves, same X or Y 139 plotting X or Y using row numbers 139 polar axes 475 quartile plots 323 range plots 321 reference lines 341 stacked bar chart 140 3D bar charts data format 109 fills 197 3D graphs adding plots 170 axes placement during rotation 389 bar chart data format 109 bar charts 375 changing view 385 creating 377, 378, 379 creating mesh 381 data format 109 examples 22, 23 generating mesh data 4 light source 385 lighting 385 line data format 109 mesh 374 mesh data format 109 mesh lines/fills 382 mesh plots 381 modifying mesh lines and fill color 382 origin axes 385 perspective 385 plotting data 147, 373 positioning axes 389 rotation 385 Scatter and line 374 scatter data format 109 smoothing mesh data 4 trajectory 378 waterfall plots 376, 379 3D mesh plots data format 109 3D mesh with contour projections 384 3D scatter plots
adding drop lines 205, 208 symbols 179, 187 95% confidence interval 709 95% confidence intervals 709 95% prediction interval 709
A
abs function 807 absolute minimum sum of squares 779, 780 accumulation functions 802 Add Procedure dialog box 524 adding axes 372 axis breaks 438 contour fills 396 contour labels 401, 402 exploding slices to pie charts 411, 412 frame lines 390 graphs to pages 243, 244 labels to page 289 macros to macros 523 multiple axes for single plot 372 plots to graphs 170 prefix/suffix to tick labels 458 procedures to macros 524 reference lines 341 regression equations to graph pages 673, 728, 746 styles to Graph Style Gallery 157 suffixes/prefixes 401 text to page 289 algorithm Marquardt-Levenberg 668, 688, 770 aligning axis titles 442 legends 299 objects/graphs/labels 2, 287 snap-to grids 289 text 289 using crosshairs 289 using snap-to 289
613 Index
with grids 288 with rulers 288 alpha value 707 alphanumeric symbols 183 angular axes about 467 arc 468 modifying 475 range 468 rotation 468 scale 468 angular values 142 ANOVA one way ANOVA transform 621 ANOVA table regression results 704 ape function 808 apex dragging to modify ternary axis ranges 480 applying data transforms 4 graph styles using the Graph Style Gallery 158 arc angular axis 468 arccos function 809 arcsin function 809 arctan function 810 area and distance functions 804 Area Below Curves 533 area beneath a curve transform 622 area function 811 area plot data formats multiple area plot 125 multiple vertical area plot 125 simple area plot 125 vertical area plot 125 area plots 343 changing area fill direction 352 changing fill color 353 climographs 354 converting multiple into complex 351
creating complex area plots 349 creating multiple and multiple vertical area plots 347 creating simple straight line area plots 344 creating vertical area plots 344 examples 310 identifying regions 351 shading between two curves 354 arguments, transform 799 arithmetic mean 64 arithmetic operators transforms 618 arranging data for 2D graphs 139 data for 3D graphs 147 data for bubble plots 144 data for contour plots 147 data for polar plots 143 graphs 286 graphs on a page 282 radial and angular values for polar plots 142 XY data for polar plots 142 arranging data 135, 146 asymmetric error bar plots 139 column averaged error bar plots 139 column means 139 polar plots 142 stacked bar chart 140 Arrhenius plot 426 Arrhenius scale 432 arrow keys moving graphs and objects using 280 arrowheads modifying for vector plots 368 arrows drawing 270 modifying arrow heads 275 ASCII files 60 aspect ratio options 236 assumption checking options 675 asymmetric error bar plots 139
614 Index
asymmetric error bars creating 2D plots with 322 asymmetrical error bars quartile plots 323 attributes 252 changing line 434 text formatting 291 automatic determination of initial parameters 766 automatic legend updating 299 automatic legends 3, 14, 293, 299 displaying 296, 298 editing 296 locking 299 restoring to default settings 296 automation 520 Add Procedure dialog box 524 adding macros to macros 523 automating routine tasks 515 creating custom dialog boxes 523 creating macros 515 creating menu commands 530 creating user-defined functions 526 Debug Window 526 editing macro code 522 editing macros 518 macro window 521 Macro Window toolbar 519 Object Browser 524 parts of macro programming language 521 running macros 532 user-defined functions 525 available statistics 64 averaging 64 avg function 811 axes displaying 433 engineering notation 455 hiding 433 modifying 433 polar plots 467 turning on/off 433 viewing 433
axis 3D placement during rotation 389 about 369 adding 372 additional for multiple plots 369 angular 475 breaks 3, 438 category scale 424, 429 changing scale 423, 424, 425, 426 changing scale types 428, 432 common log 423 contour range values 398 creating multiple for single plot 372 custom scale 432 date/time 425 dragging to modify ternary ranges 480 Extreme Value Distribution axis scale 432 labels 442 line attributes 434 linear scale 423 logit scale 424 modifying 389 modifying range by dragging 480 moving 2D 436 multiple 3 multiple pairs 372 natural log scale 423 origin axes 385 polar 475 positioning 3D 389 positioning using Graph Properties dialog 436 probability scale 423 probit scale 423 radial 475 radial tick labels 475 range values 427 range, changing 427 reciprocal 426 scale types 423, 424, 425, 426, 428, 432 scale values 458 tick labels 458 tick marks 448, 489 titles 442
615 Index
types 423, 424, 425, 426 using drawing tools, using formatting commands 4 using Object Properties dialog 434 x and y 3 axis breaks creating 438 axis range scale 427 values 427 axis scale Arrhenius 432 Extreme Value Distribution 432 switching to category 429 switching to custom 432 types 422 axis titles editing 440 hiding 441 moving 442 rotating 440 viewing 441 axis, range contour plots 398 Axon files importing 61
B
back planes color 464 grid lines 466 backup files 14 bar charts automatic reference lines 226 creating 311 edges 272 error bars 331 examples 310 fills 197, 272 grouped bars with error bars 336 histograms 229, 233 horizontal data format 109
needle 229 needle plot 202 spacing bars 202, 205, 336 step 229 bar charts, 3D creating 377, 378 examples 375 bar widths variable 205 bars edges 272 fills 197, 272 spacing 202, 205, 336 Base using as tick labels 403 base using as tick labels 455 Batch Process Excel Files 534 bidirectional error bars data format 113 bin values histograms 229 bivariate statistics transform 622 block function 812 blockheight 813 blocks, data deleting 87 inserting 85 sorting 571 blockwidth 813 Border Plots 535 box plots 10th/90th percentiles 341 box fills/color 342 box widths 342 Cleveland percentile method 342 computing percentile methods 342 computing percentile values 339 creating 311, 339 data format 109 edges 272 examples 311 fills 197, 272
616 Index
macro for creating 535 mean lines 342 modifying 341, 342 outliers 341 Standard percentile method 342 symbols 179, 187, 342 whisker cap widths 341 boxes drawing 270 edges 272 fills 197, 272 breaking links 264 breaks 438 bubble plots 179 2D 363 applications 363 arranging data 144 arranging data for size 145 converting area data to diameters 145 plotting data 144 transforms dialog box 145 X, Y values 145 By Group Data Split 536
cells engineering notation 80 formatting in worksheets 80 moving to 47 setting decimal places in worksheets 80 using as column or row titles 92 wrapping text 80 centering axis titles 442 Changing axis color 434 axis thickness 434 changing 3D graph view 385 area plot fill color 353 axis range 427 axis scale 422 axis scale types 428, 432 bar/box fills 272 color of fill pattern lines and edge lines 272 colors 194 column width 69, 71 contour labels 401, 402 contour plots 394, 402 contour range values 398 default layout template file 286 fill patterns 272 fills 194 grids 464 inserted object icons 262 line end attributes 275 line types 190, 273 multiple selected objects 277 object background fills 272 objects fills 272 page color 304 pasted object icons 261 pattern and edge line thickness 272 pattern density of plot fills 194 patterns 194 pie charts 412 polar plot axes 467 polar plots 143
C
calculating confidence intervals 224 error bar mean 331 error bars 331 linear regressions 224 prediction intervals 224 Calculation 224 cancelling a regression 690 category axis scale 424 category data 136 creating graphs using 313 using to create plots 324 category scale creating 429 modifying 429 cell 813
617 Index
radial axes 472, 475 radial axes tick labels 475 range direction 484 scale direction 484 slice fills 272 source files for links 264 symbol fills 272 ternary axis direction 484 ternary graphs 419 ternary range 484 ternary scale type 483 text formatting 291 tick label text 455 tick values 458 units of measurement 303 characters non-keyboard 289 using as symbols 183 chart fills color incrementing 197 charts creating pie charts 136 chi-square reduced 703 chisquareden function 813 chisquaredist function 814 chisquareinv function 815 Cholesky decomposition 224 choose 815 clearing 85 graph titles 269 graphs/objects 269 legends 269 Cleveland percentile method using for box plots 342 using for quartile box plots 342 climographs 354 Clipboard cutting and copying data 79, 254 using 254 clockwise angular units setting for polar plots 415 coefficient of determination
stepwise regression results 691, 702 coefficient of determination (R Squared) transform 627 coefficient of variation parameters 691 coefficients regression results 703 col 816 col function 799 color axis lines 434 bars 197 box plots 197, 342 changing 194 contour fills 396 contour lines 394 custom incrementing schemes 198 custom, using 305, 308 error bars 326 frame lines 390 graph back planes 464 grids 464 incrementing chart fills 197 line 273 lines 190 mesh lines/fills 382 page 304 radial axes 474 reference lines 227 tick marks 451, 489 color incrementing assigning to worksheet 198 customizing 198, 205 symbols 181 Color Transition Values 536 colors creating gradients 536 column averaged error bar plots 139 arranging data 139 column averaging grouped bar chart 336 column means 139 Column picker dialog box
618 Index
Graph Wizard 166 column picker dialog box normalizing ternary data 596 column statistics maximum value 64 mean 64 minimum positive value 64 minimum value 64 missing values 64 other values 64 printing 103, 104 setting Options 14 size of sample 64 standard deviation 64 standard error 64 sum of sample 64 viewing 4, 64 column titles duplicate 44 using transforms as 596 column width changing 69, 71 columns asymmetric error bar plots 139 averaging 64 column and row titles dialog box 88 column averaged error bar plots 139 deleting 87 inserting empty 85 inserting graphic cells 198, 354 inserting symbol size values 187 key 571 merging 539 multiple Z 146 picking different data for current plot 166 plotting multiple curves, same X or Y 139 plotting X or Y using row numbers 139 plotting XYZ 147 selecting 83 sizing 69, 81 sorting data 571 stacked bar chart 140 statistics 64
switching from rows to columns 87 tick labels 402, 462 titles 88, 92 type labels 462 using as row titles 90 Comma Delimited (.CSV) importing 51 command prompt using to run SigmaPlot macros 530 commands creating using macros 530 embedded graphs 177 comments entering regression 753 common logarithmic axis scale 423 completion status messages regression results 719 complex 816 complex area plots 349, 351 compression formats post-processing .TIFF files 510 Compute 1st Derivative 537 computing 224 1st derivative 537 computing percentile methods 342 cumulative percentages 545 power spectral density 544 quartiles 342 rank and percentile 545 confidence and prediction bands 673, 727, 746 confidence bands 696 nonlinear regression 678 confidence interval 709 95% 709 regression results 709 confidence intervals adding to 2D graphs 223 calculating 224 linear regressions 223 view in nonlinear regression reports 678 confidence lines 95% and 99% confidence interval 64 defined 224
619 Index
configuring printer settings 103, 105, 106 constant variance 675 P values 675 testing 675 constant variance test regression results 706 constraints entering 685 in Regression Wizard 684 parameter 684 Constraints, parameter defining 686 constraints, parameter badly formed 720 entering 768 viewing 692 constructor notation example of use 621 regression example 762 contour fills color 396 modifying 397 contour labels changing frequency 401 numeric 403 rotating 401 skipping 462 suffixes/prefixes 401 text attributes 406 turning on/off 401 contour lines color 394 line types 394 modifying 394 showing/hiding 394 thickness 394 contour plot 21 contour plots adding fills 396 adding labels 401, 402 creating 392 data format 109
editing contour labels 406 example 21 frequency of contour labels 401 line types 394 modifying 394, 402 modifying contour lines 394 modifying fills 397 modifying labels 401, 402 modifying Z data range/scale 398 plotting data 147, 392 setting line intervals 399 setting the direction of fills 396 X,Y,Z values 147 contours adding to 3D mesh plots 384 projecting onto 3D mesh plots 384 converting area data to diameters for bubble plots 145 date and time data to numbers 714 numeric and date and time data:Date and time data:converting to numeric 82 numeric data to date and time data 717 Cooks Distance 679 Cooks Distance test results 708 copy and Paste 60 Copy shortcut 48 copying data 52, 84 graphs 244 notebook items between notebooks 37, 39 objects/graphs 254, 255 correlation coefficient regression results 691, 702 cos 817 cosh 817 count 818 COUNT function 621 creating additional axes for multiple plots 369 additional plots 170 axis breaks 438 category scale 429
620 Index
complex area plots 349 contour plots 392 creating pie charts 136 custom dialog boxes 523 custom graph page layout 285 custom scale 432 embedded graphs 177 equation 35 equations to plot 208 Excel worksheets 35 files for figure submission to journals 508 filled contour plots 392 graph pages 35 graphs 150 graphs using Excel worksheets 102 graphs using the Graph Style Gallery 156, 157 graphs with Graph Toolbar 151 graphs with Graph Wizard 151 histograms 229, 233 labels 289, 291 legends 289, 291 macros 35, 515, 517 menu commands using macros 530 multiple area plots 347 multiple axes for single plot 372 multiple curves 138 new graph for current page 243 new notebook files and items 35 new object to insert 262 page templates 251, 267 pie charts 136 plots using category data 324 plots with date and time scale 431 polar plots 143, 415 reports 35 sections 35 simple straight line area plots 344 survival curves 570 ternary graphs 417 text labels, legends 289 user-defined functions 526 vector plots 570 vertical area plots 344
worksheets 35 crosshairs 289 cumulative percentages computing 545 curve fitter introduction 668 curve fitting date and time data 713 curves coefficient of determination 627 column averaged error bar plots 139 fitting date and time data 713 multiple for polar plots 142 multiple in graph 313 multiple, same X or Y 139 plotting multiple 138, 313 plotting X or Y using row numbers 139 transform for integrating under a curve 622 using category data 313 custom color 305, 308 custom dialog boxes 523 custom error bars 329 custom graph styles Graph Style Gallery 156, 158 saving 156, 158 custom scale creating 432 custom tick mark intervals 448 customer service 25 customizing color increments 198, 205 error bar directions 329 fill increments 198, 205 graph styles 156, 157 line increments 198, 205 symbol increments 198, 205 tick labels 402, 462 tick mark intervals 448 Cut shortcut 48 cutting data 84 notebook items between notebooks 37, 39 objects/graphs 254, 255
621 Index
D
data 818 2D graphs 309, 339 3D graphs 373 applying transforms 4 arranging for 2D graphs 139 arranging for 3D graphs 147 arranging for bubble plots 144 arranging for contour plots 147 arranging for polar plots 142, 143 column statistics 64 contour plots 147, 392 converting bubble plot area 145 converting data to diameters for bubble plots 145 converting date and time data to numeric data 714 converting numeric data to date and time data 717 converting to mesh format 4 copying and pasting from other applications 52 curve fitting date and time data 713 cutting 84 deleting 84, 85 entering 41, 103 entering into a worksheet 49 exporting 62 generating random data 608 highlighting outliers 95 importing 51, 52, 60, 92 inserting 50 long form mesh format 147 mesh, converting to 4 moving 85 multiple-curve plots (ternary triplets) 143 normalizing for ternary graphs 596 one column for multiple curves in polar plot 143 pasting 84 plotting additional 170 plotting different data for current plot 166 plotting portion of 173 plotting X or Y using row numbers 139
polar plots 143 previewing before printing 104 printing 103, 104 protecting data on the Web 505 radial and angular values 142 rearranging 87 regression 4 removing outlying data 92 sampling 173 saving 33 selecting 83 selecting for ternary graphs 417 single-curve plot (ternary triplets) 143 smoothing 4 smoothing 2D high-frequency data 599 smoothing 3D mesh data 604 sorting 571 ternary graphs 417 transposing 87 using transform language 608 viewing for embedded graphs 178 X,Y, many Z for contour and mesh plots 147 XY values for polar plots 142 data brushing 71, 92, 95 data feedback set colors on worksheet 71 data format 3D bar chart 109 3D graphs 109 3D line plot 109 3D mesh plot 109 3D scatter plot 109 box plots 109 contour plot 109 graph styles 113 graph types 109 horizontal bar chart 109 horizontal dot plot 113 line and scatter plot 109 line plot 109 long form mesh 147 multiple area plot 125 multiple horizontal step plot 119
622 Index
multiple regressions 113 multiple scatter plot 113 multiple spline curves 119 multiple straight line 122 multiple vertical area plot 125 multiple vertical step plot 119 pie chart 109 polar plot 109 scatter plot 109 simple area plot 125 simple scatter plot 113 simple spline curve 122 simple straight line 122 simple vertical plot 119 ternary 109 vertical area plot 125 vertical bar chart 109 vertical dot plot 113 data format options Regression Wizard 682 data formats 135, 146 pie charts 136 polar plots 142 data manipulation functions 801 databases importing 53 date and time Options 14 tick intervals 447 date and time axis scale tick labels 459 tick labels for contour plots 404 tick marks 447 date and time data converting to numeric data 714 curve fitting 713 formatting in worksheets 80 date and time format date delimiters 77 entering 50 regional settings 77 using with Excel 79 worksheet display 75, 79
date and time scales creating plots with 431 date delimiters date and time formats 77 date/time axis scale 425 Day Zero leap years 77 setting in worksheets 77 dBase (.DBF) importing 51 Debug Window 526 Intermediate tab 527 Stack tab 528 tabs 527 toolbar buttons 526 Watch tab 527 decimal places setting in worksheets 71 defaults setting graph 108 defining 763 degrees 596 degrees of freedom regression results 704 Delete Cells... command 48 Delete shortcut 48 deleting 84 columns and rows 87 data 84, 85 graph titles 269 legends 269 objects 269 density changing pattern of plot fills 194 dependencies exponential equation 785 parameter 691, 785 dependent variables entering 755 derivatives computing 537 descriptions of transform functions 800
623 Index
design graphing references 27 principles of graphing 27 suggested reading 27 determining initial parameters 766 DFFITS 679 DFFITS test regression results 708 diagnostics influence 708 regression results 707 dialog boxes creating in SigmaPlot 523 diff 819 DIFF function 622 differential equation solving 623 digital pre-press preparing graphs 507 direction customized error bars 329 error bars 327 reference lines 227 display formats date and time format 75, 79 numbers 72 displaying automatic legends 296, 298 axes 433 contour fills 396 contour labels 401, 402 contour lines 394 grid lines 466 outlying data 95 page margins 301 reference lines 227 dist 819 distance functions 804 dividing data into multiple columns 536 docking
Graph Style Gallery 157 Notebook Manager 31 DOS files importing 51 dot plot horizontal data format 113 vertical data format 113 dpi 510 dragging 2D axes 435 Notebook Manager 31 objects 277 radial axes 472 ternary axes 480 drawing arrows 270 ellipses 270 lines 270 objects 4, 270 Page toolbar 270 drawing speed 173 drop lines adding to plots 205, 208 attributes 205 for single point 208 modifying 205 dsinp 820 duplicate column titles 44 Durbin-Watson 675 dynamic curve fitting 721 Dynamic Fit Wizard 721, 722 creating new equations 681 Equation Library 723 equation options 683 examples 730 finishing the fit 728 graphs 730 reports 728 selecting data 722 selecting the equation 723 selecting variables 723 setting curve fit options 724 setting graph options 727
624 Index
setting numeric output options 725 setting results options 726 using to add equations to graph pages 728 viewing and editing code 681 viewing initial results 725
E
E notation 67 E Notation Always display 72 E Notation When Needed display 72 edge lines 194 setting the color 272 setting thickness 272 edges bar/box 272 plot line thickness 194 slice 272 Edit menu commands 84 Copy 84 Cut 84 Delete Cells... 85 Insert New Object 262 paste 84 Transpose Paste 87 editing automatic legends 296 axis tick labels 453 axis titles 440, 442 contour labels 401, 406 embedded graphs 177, 178 equations 681 Graph Style Gallery graphs 159 graph titles 166 macros 518, 522 notebook items 37 notebook sections 37 object links 264 page format 300 pasted graphs in other applications using OLE2 255, 264
plot name 164 radial axes tick labels 475 text 291 tick labels 455 ellipses drawing 270 Embedded graphs resizing 178 embedded graphs available menus and command 177 editing 177, 178 opening inside SigmaPlot 178 viewind data 178 viewing data 178 embedding objects 257, 266 Reports 257 embedding objects in graphs identifying objects on page 259 viewing as an icon 258 engineering notation 72 as used in SigmaPlot 67 scientific notation 67 setting in worksheets 80 tick labels 455 using as axis values 455 entering column titles 88, 92 constraints 685 constraints, parameter 768 data 41, 103 data into worksheets 49 equations 208 equations, regression 754 Greek symbols 289 iterations 687, 769 labels 289 options 769 parameters 765 regression comments 753 regression equation settings 752 regression statements 754 row titles 88, 92
625 Index
step size 688, 770 symbols in legends 289 text 289, 291 tolerance 688, 770 variables 755 EPS files 509 equation curves extending to axes 673, 727, 746 equation solving 789 equations adding to graph pages 673, 728, 746 creating 208 creating within Notebook Manager 35 editing 681 linear regression 224 manually entering 208 overparameterized 794 plotting 208, 218 prediction intervals 226 saving 758 setting options in the Dynamic Fit Wizard 724 setting options in the Global Fit Wizard 744 setting options in the Regression Wizard 672 setting parameters 208, 213 solving 216 solving guidelines 218 equations, regression entering 754 fit statements 774 iterations 687 logistic 789 parameters 765 regression statements 720, 754 results 690 results messages 719 running again 692 saving results 693 solving 789 step size 688, 770 tolerance 688, 770 variables 755 weight variables 760 equations, transform
variables 619 error bar direction 327, 329 error bar plots arranging data 139 error bars asymmetric 322 bidirectional data format 113 bidirectional error bars 113 calculating 331 cap width 326 color 326 creating 2D plots with 317 creating grouped bar charts with 333 custom directions 329 direction 327 generating 331 grouped bar charts 336 horizontal data format 113 line thickness 326 mean computation method 331 methods for generating 332 modifying 325 modifying appearance 326 multiple and regressions data format 113 multiple data format 113 plot types 317 quartile plots 323 range plots 321 relative direction 327 simple and regression data format 113 simple data format 113 vertical point plot 113 error status messages regression results 720 European address and phone number 25 evaluating F at 216 mathematical expresions 218 mathematical expressions 216 examples 21 2D graphs 15, 21 3D graphs 22, 23 dynamic curve fitting 730
626 Index
of macro uses 528 pie charts 15 polar plot 20 transforms 620 Excel importing 58 Excel files batch processing 534 Excel workbooks options 14 Excel worksheets 102 creating graphs 102 creating within Notebook Manager 35 limitations 99 opening 96 opening data files 99 regression 102 statistics 99 system requirements 6 toolbars 101 transforms 102 unprotecting workbooks 98 using 96, 103 using date and time format 79 using Excel print commands 100 workbooks 4 executing one-line functions 591 exp 820 exploding pie chart slices 411, 412 exponential equations dependency example 785 exponents numeric tick labels 403, 455 exporting data associated with graph 506 data only, not the graph 506 Excel worksheets 100 graphs and pages 237 graphs as webpages 506 inserting graphs into FrontPage 507 into HTML 504 reports 496
to Systat 63 worksheet data 62, 103 worksheets 62 worksheets as text files 63 extended transforms 591 extending equation curves to axes 673, 727, 746 Extreme Value Distribution scale 432
F
F statistic regression results 704 factor axis tick labels 455 factorial 821 FAQs 25 fast Fourier functions 806 fast page open 14 fden function 822 fdist function 822 feedback mouse-over 242 fft 823 file formats 51 File menu commands Page Setup 301 file types 99 files as objects to insert 262 breaking links between source and object 264 changing source for linked objects 264 embedding objects 257, 266 exporting as non-notebook files 237 importing data from 51, 60, 92 linking objects 257, 266 notebook templates 251, 267 opening non-notebook 38, 39 saving 33 saving notebook 33 text 60 updating links 264 fill color
627 Index
modifying 382 filled contour plots creating 392 modifying 397 fills area plots 352, 353 bar chart 272 box plot 272 box plots 342 change color of pattern lines and edge lines 272 change object fill background color 272 change object fill pattern 272 changing 194 contour plots 397 custom incrementing schemes 198 increment customizing 198, 205 mesh plots 382 modifying 197 object 272 pie chart 272 symbol 272 finv function 824 fit f to y with weight w 760 fit statements modifying 774 fit with weight 686 Fixed Decimal display 72 fonts Greek 289 PostScript 289 symbols 289 TrueType 289 for 824 Format menu commands Align 287 line 273 formats submitting graphs for publication 507 formatting cells 80 date and time data 80 date and time tick labels 459
labels 3 numeric data in worksheets 80 text 291 text in cells 80 fractional defective control chart transform 628 frame lines color 390 line type 390 modifying 390 relative to origin 390 relative to viewer 390 freezing panes 46 frequency contour labels 401 frequency plots creating 537 frequently asked questions 25 FrontPage inserting graphs into 507 function arguments 799 Function dialog box 681 functions abs 807 accumulation 802 ape 808 arccos 809 arcsin 809 arctan 810 area 811 area and distance 804 avg 811 block 812 blockheight 813 blockwidth 813 cell 813 chisquareden 813 chisquaredist 814 chisquareinv 815 choose 815 col 816 colL 799 complex 816 cos 817
628 Index
cosh 817 count 621, 818 curve fitting 804 data 818 data manipulation 801 descriptions 800 diff 622, 819 dist 819 distance 804 dsinp 820 exp 820 factorial 821 fast Fourier 806 fden 822 fdist 822 fft 823 finv 824 for 824 fwhm 826 gaussian 827 histogram 827 IF 620 if 621, 828 if...then...else 829 imaginary (img) 831 int 831 interpolate 832 inv 832 invcpx 833 invfft 834 ln 834 log 835 logistic 789 lookup 835 lowess 837 lowpass 838 max 839 mean 621, 622, 839 min 840 miscellaneous 805 missing 840 mod 841 mulcpx 841
normden 842 normdist 842 norminv 843 nth 844 numeric 802 one-line 591 partdist 844 polynomial 845 prec 846 precision 803 put into 846 random 847 random number 803 range 802 real 848 regression 791 rgbcolor 848 round 849 runavg 849 sin 850 sinh 851 sinp 851 size 852 solving 216, 218 sort 852 special constructs 805 sqrt 628, 853 statistical 803 stddev 622, 628, 854 stderr 854 subblock 855 sum 855 tan 856 tanh 856 tden 857 tdist 857 tinv 858 total 621, 622, 859 transforms 799 trigonometric 801 worksheet 800 x25 859 x50 860
629 Index
G
galleries graph style 156 gap colors in line 190 gaussian 827 Gaussian Cumulative Distribution 538 generating error bars 331 linear regression lines 220, 223 mesh data 4 global changing multiple page objects 277 global curve fitting 742 Global Fit Wizard 742 creating new equations 681 Equation Library 743 equation options 683 finishing the fit 746 reports 746 selecting data 743 selecting parameters to share 744 selecting the equation 743 selecting variables 744 setting graph options 745 setting numeric output options 745 using to add equations to graph pages 746 viewing and editing code 681 viewing initial results 745 global text changes 291 Go to... worksheet cell 47 goolbars graph 151 gradient colors creating 536 grads 596 graph defaults 3, 14
options 108 setting 108 graph dialog boxes Graph Properties 239 Graph Wizard 239 graph page setting options 236 graph pages creating within Notebook Manager 35 naming 36 printing 238 selecting objects 240 Graph Properties dialog box 239 customizing tick labels 462 modifying graphs 162 modifying grids and planes 162 modifying tick appearance 442, 449 modifying titles and legends 162 moving axes 436 pie charts 136 positioning 2D axes 436 radial axes 472 viewing/hiding axis 433 graph style custom 156 Graph Style Gallery 156 multiple straight lines 119 saving 156 simple straight line 119 graph style data format Horizontal dot plot 113 horizontal point plot 113 hortizontal error bars 113 multiple area plot 125 multiple error bars 113 multiple error bars and regressions 113 multiple horizontal step plot 119 multiple regressions 113 multiple scatter plot 113 multiple spline curves 119 multiple straight line 122 multiple vertical area plot 125 multiple vertical step plot 119
630 Index
simple area plot 125 simple error bars 113 simple error bars and regression 113 simple horizontal step plot 119 simple regression 113 simple scatter plot 113 simple spline curve 119, 122 simple straight line 122 simple vertical step plot 119 vertical area plot 125 vertical dot plot 113 Graph Style Gallery adding styles 157 applying graph styles 158 creating graphs 156, 157 docking graphs 157 editing graphs 159 moving 157 using 156 graph styles custom graph styles 158 Graph Style Gallery 158 using the Graph Style Gallery 156, 157 graph titles deleting 269 Graph Toolbar 151 graph toolbars 2D, 3D 243 graph type data format 3D bar chart 109 3D line plot 109 3D mesh plot 109 3D scatter plot 109 box plots 109 contour plot 109 horizontal bar chart 109 line and scatter plot 109 line plot 109 pie chart 109 polar plot 109 scatter plot 109 ternary 109 vertical bar chart 109
Graph Wizard additional axes 369 area plots 344 bubble plots 363 changing graph type/style 168 creating 3D graphs 377, 378, 379 creating graphs 151 creating mesh plots 381 multiple plots 369 vector plots 365, 366 graphs 2D example 15 3D example 22, 23 adding drop lines 205, 208 adding plots 170 adding to page 243, 244 aligning 287 anatomy of 15 arranging on a page 282, 286 asymmetric error bar plots 139 automatic legends 293, 299 axes 389 box plots 341 centering 287 column averaged error bar plots 139 contour plots 21, 409 copying 254, 255 creating 150 cutting 254, 255 data format for graph styles 113 data format for graph types 109 defaults, setting 108 displaying automatic legends 296, 298 editing in Graph Style Gallery 159 editing in other applications using OLE2 255, 264 grid lines 466 grouped bar charts 333, 334 grouping/ungrouping objects/text 281 hiding 268 hiding automatic legend 296, 298 hiding graph titles 269 hiding on page 267, 268
631 Index
hiding using shortcut menu 268 legends 289 modifying 2D plots 220, 227 modifying type/style 168 modifying using Graph Properties dialog 162 moving 277, 279 multiple curves, same X or Y 139 naming plots 164 pasting without data 259 picking different data for current plot 166 pie charts 15, 409 plotting data 139 plotting multiple curves 138 plotting X or Y using row numbers 139 polar plots 20, 409 references for design 27 resizing for publication 510 resizing labels/legends automatically 236 saving 33 scaling 277, 279 scatter, 3D 147 selecting 163 selecting style 113, 150 selecting type 109 sizing 277, 279 stacked bar chart 140 styles 113 symbols 179, 187 ternary 409 titles 166 types 109 types and styles 1 using Paste Special 255, 259 viewing on page 267, 268 working on pages 266, 267, 272, 277, 305, 308 zooming in/out 245 graphs, 2D 311 Greek symbols entering 289 Gregorian calender 77 grid changing colors 70 grids
aligning graphs and objects 288 color 465 displaying in front/behind 465 graph back planes 466 hiding 466 line types 465 mesh graphs 4 modifying 389 snap-to 289 turning on/off 466 grouped bar charts column averages 336 creating 333, 334 creating with error bars 333 error bars 336 examples 333 spacing bars 202, 205 grouped data 313 grouping objects/text 281 guidelines equation solving 218 for submitting graphs for publication 507
H
halting 299 hardware system requirements 6 Help system using 24 hiding automatic legends 296, 298 axes 433 axis titles 441 contour lines 394 graph titles 269 graphs on page 267, 268 graphs using shortcut menu 268 grid lines 466 legends 269 notebooks in Notebook Manager 31 radial axis labels 475
632 Index
statistics 66 tick marks 441 high-frequency data smoothing 599 highlighting outliers 95 histogram 827 Histogram Wizard using 229 histograms bin values 229 creating 229, 233 histogram transform function 229 Histogram Wizard 229, 233 histogram.xfm transform 233 macro for creating 535 homoscedasticity constant variance test 706 horizontal bar chart data format 109 horizontal dot plot data format 113 horizontal error bars data format 113 horizontal point plot data format 113 HTML exporting graphs into 504
I
icons changing display for inserted objects 262 changing display for pasted objects 261 displaying inserted objects as 262 displaying pasted objects as 261 on graph pages 258 identifying area plot intersections 351 if 828 IF function 621 logical operators 620 if...then...else 829
ignoring outliers 92 imaginary (img) 831 Importing SigmaPlot files 58 SigmaScan files 58 SigmaStat files 58 importing Axon files 61 data by copying and pasting 52 data files 51, 60, 92 data from multiple Excel files 534 databases 53 Excel files 58 Lotus 123 files 59 MicroSoft Excel files 59 ODBC databases 53 Quattro files 59 SPSS files 61 text files 60 inches page units 303 incrementing lines 192 symbol color 181 incrementing colors chart fills 197 incrementing line types 190 incrementing schemes customized 198 independent graph pages 37 independent variables entering 755 indexed data 136, 313 influence 679 influence diagnostics regression results 708 influential point tests 708 Insert Cells shortcut 48 Insert Date and Time command for reports 502 inserting columns and rows 85
633 Index
data 50 displaying inserted objects as icons 262 graphs into FrontPage 507 graphs into Microsoft Word 538 linked objects 262 modifying inserted object icons 262 new object 262 objects from file 262 Insertion mode turning on/off 50 installing SigmaPlot 5 serial numbers 6 types of folders 6 int 831 integrating under curve transform 622 interpolate 832 Interpolating data setting mesh range values 604 interpreting results regression 690, 693 intersections identifying in area plots 351 intervals confidence/prediction 64, 223 setting for contour plots 399 tick mark values assigned to a worksheet 448 inv 832 invcpx 833 invfft 834 iterations convergence 668 entering 687, 769 exceed maximum numbers 719 more iterations 719
key column 571 keyboard moving around worksheet 47 moving graphs and objects using arrow keys 280 keystrokes functions 47 Kolmogorov-Smirnov test 675
L
label notation 455 labeling with symbols, text from column 539 labels adding to page 289 aligning 3, 287 automatic scaling with graphs 236 axis titles 442 axis values 458 column titles 88, 92 column type 462 contour 401, 402 creating 289, 291 editing tick 455 entering non-keyboard characters 289 formatting 3 frequency of contour 401 graph titles 166 grouping/ungrouping 281 numeric tick 402 radial axes 475 reference lines 227 rotating 289 rotating contour 401 row titles 88, 92 suffixes/prefixes 401 tick mark 458 using column and row title dialog box 88 using for column titles 89 using for row titles 90 landscape page orientation 302
J
JNT files 282 journals preparing graphs for publication 510
K
Kaplan-Meier survival curve 570
634 Index
layering in front/behind grid lines 465 reference lines 227 layout graph design references 27 layout files 14 leap years 77 least squares regression 781 legends 269 adding symbols 289 adding to page 289 aligning 299 automatic 3, 293, 299 automatic scaling with graphs 236 creating 289, 291 deleting 269 editing 296 hiding 296 locking 299 restoring to default settings 296 showing 296 ungrouping 299 leverage 679 leverage test regression results 708 light source 3D graphs 385 line and scatter plot multiple straight line data format 122 simple straight line data format 122 line and scatter plots adding drop lines 205, 208 creating 311 data format 109 examples 310 line plot data formats multiple horizontal step plot 119 multiple spline curves 119 multiple vertical step plot 119 simple horizontal step plot 119 simple spline curve 119 vertical step plot 119
line plots 3D data format 109 adding drop lines 208 adding drop lines to 205, 208 color 190 creating 311 data format 109 examples 310 midpoint step 190 multiple straight lines 119 step 190 straight 190 symbols 179, 187 line type frame lines 390 modifying 273 reference lines 227 line/scatter graphs asymmetric error bars 322 error bars 317 quartile plots 323 range plots 321 symbols 179, 187 linear axis scale 423, 455 linear interpolation for computing percentile methods 342 linear regression dialog box parameter values transform 627 standard deviation 627 linear regressions calculating 224 confidence/prediction intervals 223 defined 224 generating 220, 223 multiple curves 220 polynomial order 220 results 222 lines adding arrow heads 275 alternating colors in lines 190 assigning to worksheet 198 attributes 434 axis 434
635 Index
changing end attributes 275 changing thickness 190 changing type 190 color 190 contour 394 custom incrementing schemes 198 drawing 270 drop 205, 208 error bars 326 frame 390 grid 466 grid line types 465 increment customizing 198, 205 incrementing 190, 192 layering in front/behind symbols 190 linear regression 220, 223 mesh plots 382 midpoint step plots 190 modifying properties 273 radial axes 472, 475 reference 341 setting intervals for contour plots 399 smoothed 190 spline curves 190 step plots 190 type 190 linking objects 257, 266 Reports 257 links breaking 264 changing source files 264 editing 264 manual/automatic updating 264 viewing object links 264, 266 ln 834 local maximum sum of squares 779 local minimum finding 777 locking legends 299 log 835
log axes 455 logarithmic axis scale 423, 455 logical operators transforms 620 logistic function four parameter 789 logit axis scale 424 lookup 835 Lorentzian distribution regression example 773 Lotus 1-2-3 (WK*) importing 51 lournals preparing graphs for publication 507 lowess 837 lowpass 838 LZW compression algorithm post-processing .TIFF files 510
M
macro language creating macros using 517 macro programming language 521 Macro Window 520 opening 517 options 521 macro window appearance 521 Macro Window toolbar 519 macros 520, 533, 534, 535, 536, 537, 538, 539, 540, 541, 544, 545, 546, 570 adding macros to macros 523 adding procedures 524 creating 515, 517 creating as menu commands 530 creating custom dialog boxes 523 creating user-defined functions 526 creating within Notebook Manager 35 editing 518, 522 examples 528 for Microsoft Word/Excel 528 included with SigmaPlot 531
636 Index
programming language 521 recording 515 running 532 running from a command line 530 setting options 521 user-defined functions 525 using Macro Window toolbar 519 using the Add Procedure dialog box 524 using the Debug Window 526 using the Object Browser 524 using to automate routine tasks 515 window options 521 magnifying page view 245 magnitude data plotting as vector 570 major ticks date and time axis scale 447 Marquardt-Levenberg Algorithm references 669 Marquardt-Levenberg algorithm 668, 688, 770 max 839 maximum value (max) column statistics 64 mean 839 column statistics 64 error bar computation method 331 mean computation method 331 MEAN function 621, 622 mean lines 341 box plots 342 mean squares regression results 704 median lines 341 menu commands creating using macros 530 menus creating menu commands using macros 530 using with embedded graphs 177 merging columns 539 mesh data converting 4
generating 4 interpolating 4 mesh lines modifying 382 mesh plots 3D data format 109 adding contours 384 creating 381 examples 374 fills/color 382 light source 385 mesh data 4 mesh lines 382 modifying lines/fills 382 plotting data 147 projecting contours 384 smoothing data 4 transparent 381 X,Y,Z values 147 messages completion status 719 error status 720 regression results 719 regression status 690 metafiles 256 Microsoft Excel opening SigmaPlot within 528 Microsoft Excel (.XLS) importing 51 Microsoft Word inserting graphs 538 opening SigmaPlot within 528 millimeters page units 303 min 840 minimum positive value (min pos) column statistics 64 minimum value (min) column statistics 64 Minitab importing 51 minor ticks date and time axes 447
637 Index
modifying 447 missing 840 missing values column statistics 64 mistakes undoing mistakes 15 Mocha Worksheets importing 51 Mocha worksheets 58 mod 841 modifying 2D plots 220, 227 3D graph view 385 adding plots 170 area plots 352, 353 attributes for new pages 252 automatic legends 293, 299 axes 389, 433 axis range by dragging 480 axis scale 422 axis scale types 428, 432 background colors 197 box fills/color 342 box plots 341 box widths 342 category scales 429 contour labels 401, 402 contour lines 394 contour plots 394, 402 drawn objects 272, 277 drop lines 205, 208 edges 194 embedded graphs 177, 178 error bar appearance 326 error bar computation method 331 error bar direction 327 error bars 325 fill color 382 fills 197 frame lines 390 graph lighting 385 graph perspective 385 graph rotation 385
graph styles 168 graph titles 166 graph types 168 graphs using Graph Properties dialog 162 grid lines 389 grids 464 grids and planes 162 inserted object ico?s 262 line color 273 line end attributes 275 line thickness 273 line type 273 mesh lines 382 mesh plots 382 modifying 197 multiple selected objects 277 multiple text labels 291 object fills 272 object links 264, 266 page color 304 page view 245 pasted object icon 261 patterns 197 pie charts 412 plot name 164 plot pattern line thickness 194 polar axes 475 polar plot axes 467 polar plots 143 radial axes 475 radial axes arc 472, 475 radial axes tick labels 475 source files for links 264 symbol attributes 179 symbols 179, 187 ternary axis direction 484 ternary plots 419 ternary tick labels 491 ternary tick marks 489 text formatting 291 tick labels 458 tick marks 389 titles and legends 162
638 Index
whisker cap widths 341 modifying for new pages 252 mouse moving objects 277 sizing objects 278 mouse-over feedback 242 moving 2D axes 436 2D axes manually 435 2D axes with mouse 435 around the worksheet 47 axes to precise location 436 axes with Graph Properties dialog 436 axis titles 442 data 85 graphs 277, 279 graphs and objects using arrow keys 280 notebook items between notebooks 37, 39 objects 277, 279 objects to back/front 281 radial axis 472, 475 to worksheet cell 47 moving objects 277 mulcpx 841 multiple 15 multiple area plots creating 347 multiple axes creating 372 multiple curves creating 313 plotting data 313 regression options 220 using category data 313 multiple error bars data format 113 multiple error bars and regressions data format 113 multiple independent variables 682 multiple plots 170 additional axes 369 multiple regressions data format 113
multiple scatter plot data format 113 multiple spline curve format 119 multiple straight line data format 122 multiple users 6 multiple Z columns 146
N
naming graph pages 36, 37 graphs 166 notebook files 36, 37 notebook items 37 plots 164 sections 36, 37 worksheets 36, 37 natural log axis scale 423 needle plot creating 202 new equation 35 Excel worksheets 35 graph pages 35 graphs with templates 247 macros 35 notebook files and items 35 pages with templates 247 reports 35 sections 35 worksheets 35 new features 5 noisy data smoothing 599 non-keyboard characters 289 nonlinear regression 627 assumption checking 675 confidence intervals 678 constant variance 675 Cooks Distance 679 DFFITS 679
639 Index
Durbin-Watson 675 influence 679 Kolmogorov-Smirnov test 675 leverage 679 normality 675 power 679 prediction intervals 678 PRESS Prediction Error 679 setting options 672, 726, 745 view in regression reports 676 nonlinear regression reports 675 confidence intervals 678 include confidence intervals 678 predicted values 676 prediction intervals 678 raw residuals 676 report flagged values only 676 Studentized deleted residuals 676 Studentized residuals 676 non-notebook files exporting to original file format 237 opening 38, 39 norm effect of weighting 764 normal using as source template for new pages 252 normality 675 P values 675 testing 675 normality test regression 706 normalize 596 normden function 842 normdist function 842 norminv function 843 notebook files creating 35 naming 36, 37 opening 38, 39 saving 33 template notebook files 251, 267 viewing 38, 39 notebook items 37
creating within Notebook Manager 35 cutting/copying between notebooks 37, 39 exporting as non-notebook files 237 naming 36 opening 38, 39 printing selected notebook items 33 saving 33 viewing 38, 39 Notebook Manager cutting/copying between notebooks 39 docking 31 dragging 31 opening and closing notebooks 31 overview 29 sizing 31 notebooks closing 31 opening 31 password protecting 34 using the Notebook Manager 29 viewing 31 novice prompting 14 nth 844 nudging graphs and objects 280 numbers display formats 72 functions 802 options 14 precision functions 803 random generation functions 803 numeric axis values factoring out 455 numeric data converting to date and time data 82, 717 formatting in worksheets 80 setting decimal places in worksheets 80 setting engineering notation in worksheets 80 numeric functions 802 numeric values changing contour labels 403 tick labels 403
640 Index
O
Object Browser 524 Object Properties dialog program options 236 specifying size 279 Object Properties dialog box axis attributes 434 modifying line 273 specifying location 279 objects aligning 4, 287 breaking links 264 change background color 272 change color of fill pattern lines and edge lines 272 change fill patterns 272 changing source files for links 264 copying 254, 255 cutting 254, 255 displaying as icons 261, 262 dragging 277 drawing 4, 270 editing linked 257, 266 editing links 264 embedding 257, 266 fills 272 grouping/ungrouping 281 identifying on page 259 inserting from file 262 inserting linked objects 262 inserting new 262 linking 257, 266 modifying 272, 277 modifying object links 264, 266 mouse, using to size 278 moving 277, 279 moving front/back 281 multiple selection 277 pasting as linked/embedded 261 pasting as specified file type 261 pasting to a page 255, 267 scaling 277, 279
selecting on a page 240 set pattern and edge line thickness 272 sizing 277, 279 updating links 264 using the Paste Special command 255, 259 viewing object links 264, 266 working on pages 266, 267, 272, 277, 305, 308 ODBC databases 53 importing Excel 58 OLE viewing objects as icons 258 OLE2 embedding 255, 264 linking 255, 264 pasting graphs 255, 264 one way analysis of variance (ANOVA) transform 621 one-line functions executing 591 opening data files into Excel worksheets 99 embedded graphs 178 Excel worksheets 96 non-notebook files 38, 39 notebook files 38, 39 notebook items 38, 39 notebooks 31 SigmaPlot as a command line 530 SigmaPlot in other applications 528 worksheets 39 opening worksheets 42 operators transform operators 616 options automatic 236 automatic legends 14 backup files 14 column statistics 14, 66 Excel workbooks 14 fast page open 14 for assumption checking 675 graph defaults 14, 108 graph pages 236
641 Index
grid colors 236 grid density 236 grids 236 layout files 14 macro window 521 nonlinear regression reports 675, 676 novice prompting 14 numbers 14 page 14 page undo 236 page units 236 residuals 676 retain window states 14 set auto recovery 14 setting date and time 14 setting program 236 show rulers 236 Snap-to 236 statistics 14 stretch maintains aspect ratio 236 templates 14 worksheet 14 Options button Regression Wizard 681 options, regression 687 step size 688, 770 tolerance 688, 770, 774 origin axes 385 other values column statistics 64 outliers box plots 341 highlighting 95 removing 92 Overwrite mode 50
P
P value regression results 704 P values for normality and constant variance 675 page adding graphs 243, 244
aligning objects/graphs 287 automatic legends 293, 299 changing format 300 clearing graphs/objects 269 color 304 copying graphs 244 copying objects/graphs 254, 255 creating graphs for current 243 creating new objects to insert 262 cutting objects/graphs 254, 255 deleting objects 269 editing format 300 embedding objects 257, 266 exporting as non-notebook file 237 hiding graph titles 269 hiding graphs 267, 268 hiding legends 269 inserting objects from files 262 legends 289 linking objects 257, 266 magnifying 245 moving between notebooks 37, 39 moving objects/graphs 277, 279 naming 36, 37 paper size 302 pasting graphs 244 pasting graphs/objects 255, 259 setting grid color 236 setting grid density 236 setting Options 14 setting options 236 setting snap-to grids 236 setting units of measurement 236 setup 300 showing grids 236 showing rulers 236 sizing graphs and objects 277, 279 specifying graph and object location 279 specifying graph and object size 279 templates 251, 267 text and labels 289 units of measurement 303 viewing full 245
642 Index
viewing graphs 267, 268 working with graphs 266, 267, 272, 277, 305, 308 working with objects 266, 267, 272, 277, 305, 308 zooming in/out 245 page objects selecting on a page 240 Page toolbar 270 Page undo options 236 pages selecting graphs 240 selecting objects 240 selecting text 240 paper size 302 parameters coefficient of variation 691 constraints 683, 768 convergence message 719 default settings in Regression Wizard 683 defined but not referenced 720 dependencies 691, 785 determining initial values 766 entering 683, 765 identifiability 794 initial values 765 invalid 719 missing 720 regression results 691 setting in equations 213 standard error 691 viewing constraints 692 parametric equations plotting 544 partdist 844 password protecting data 505 passwords password protecting 505 protecting notebooks 34 Paste shortcut 48 Paste Special command 255, 259
displaying pasted objects as icons 261 embedding objects 257, 266 linking objects 257, 266 modifying pasted object icons 261 pasting graphs without data 259 pasting data 52, 84 graphs as metafiles 256 graphs to page 244 graphs to PowerPoint 540 graphs/objects with the Paste Special command 255, 259 objects 255, 259 transpose 87 patterns assigning to worksheet 198 changing 194 changing color lines of fills and edges 272 changing density of plot fills 194 changing object fills 272 modifying 197 plot line thickness 194 set pattern and edge line thickness 272 setting thickness 272 percentile methods box plots 342 percentiles 342 performing extended transforms 591 perspective 3D graphs 385 Photoshop post-processing .TIFF files 510 picking data different columns to plot 166 different data for current plot 166 Pie charts sample graph 19 pie charts 197 adding exploding slices 411, 412 creating 136 data format 109 data formats 136
643 Index
example 15 examples 409 fills 197, 272 modifying 412 plotting data 136 rotating 411 slice edges 272 Piper plots 541 Plain Text (.TXT, .PRN,.DAT,.ASC) importing 51 plot fills pattern density 194 plot styles multiple curves 313 plot types error bars 317 plots about 369 adding new 170 contour 409 custom incrementing schemes 198 dummy 372 frequency 537 layering lines in front/behind symbols 190 multiple 170 multiple axes for single 372 naming 164 needle plot 202 offsetting radial axes 472 pattern line thickness 194 picking different data 166 pie charts 409 Piper plots 541 polar 409 polar axes 475 selecting 163 survival curve 570 symbols 179, 187 ternary 409 types 1 vector plots 570 plots, 2D box plots 341
creating with asymmetric error bars 322 creating with error bars 317 grouped bar charts 333, 334 line 310 line and scatter 310 linear regression lines 220, 223 modifying 2D 220, 227 multiple curves 313 reference lines 341 scatter 310 types available 310, 311, 339 using category data 313 plots, 3D bar charts 375 mesh 374 scatter and line plots 374 types available 373 waterfall plots 376 plotting equations onto existing graphs 211 magnitude data as vector 570 polar and parametric equations 544 portion of data 173 saved equations 214 plotting data 2D graphs 139 3D graphs 147 additional data 170 asymmetric error bar plots 139 bubble plots 144 column averaged error bar plots 139 contour plots 147 multiple curves 138, 313 multiple curves, same X or Y 139 polar plots 143, 415 portion of 173 scatter graphs, 3D 147 stacked bar chart 140 using category data 313 using row numbers for X or Y values 139 plotting equations 208, 218 point plot horizontal data format 113
644 Index
vertical data format 113 points page units 303 polar and parametric equations plotting 544 polar axes modifying 475 radial tick labels 475 polar plots angular axes 468 arranging data 142, 143 clockwise angular units 415 creating 143, 415 data for multiple curves 142 data format 109 example 20 modifying 143 modifying axes 467 offsetting radial axes 472 plotting data 143 radial and angular values 142 radial axes 472, 475 symbols 179, 187 using XY values 142 polar plots, multiple curves data from one column 143 polynomial 845 polynomial order regression lines 220 population confidence interval results 709 portrait page orientation 302 positioning 2D axes using Graph Properties dialog 436 2D axis 436 3D axes 389 power 679, 707 alpha value 707 regression results 707 power spectral density computing 544 PowerPoint
creating slides with graphs 540 prec 846 precise location moving axes to 436 precision functions 803 precision options 455 predicted values regression diagnostic results 707 regression results 708 view in nonlinear regression reports 677 view in regression reports 676 prediction bands 696 nonlinear regression 678 prediction interval 709 95% 709 regression results 709 prediction intervals adding to 2D graphs 223 calculating 224 defined 226 linear regressions 223 view in nonlinear regression reports 678 prefixes contour labels 401 tick labels 458 preparing graphs for publication 507, 510 PRESS Prediction Error 679 previewing worksheets before printing 104 printer settings configuring 103, 105, 106 printing column statistics 103, 104 Excel worksheets 100 graph pages 238 guidelines for submitting graphs for publication 507 previewing 104 reports 497 selected notebook items 33 setting options 103, 105 worksheet 103, 104
645 Index
worksheet data 103, 104 printing options setting 238 probability axis scale 423 probit axis scale 423 procedures adding to macros 524 producing file for publication 508 producing files for publication .EPS files 508 .TIFF files 508 SigmaPlot files 508 program folders 6 projecting contours onto 3D mesh plots 384 properties modifying text 291 protecting notebooks with passwords 34 publishing graphs in journals 507, 510 journal submission requirements 507 on the World Wide Web 503 publishing graphs 504 .EPS 509 about dpi 510 tips and tricks 510 using the Submission Assistant 512 put into 846
R
radial axes about 471 attributes 474 lines 472, 475 modifying 472, 475 moving 472, 475 offset from graph center 472 tick labels 475 turning on/off 472, 475 radial labels turning on/off 475 radial values 142 radians 596 random 847 random generation functions 803 range angular axis 468 Z data for contour plots 398 range plots 321 range, axis axis values 398 modifying 427 modifying by dragging 480 ranges functions 802 operators 618 rank and percentile computing 545 raw residuals view in regression reports 676 real 848 rearranging data 87 reciprocal axis scale 426 recording macros 515 redo 15 reduced chi-square 703 reference lines adding to 2D graphs 341
Q
quality control lines 341 quartile plots 323 quartiles computing for box plots 342 Quattro/DOS (.WK*) importing 51 quick re-plot 545 Quick Transforms using as column titles 596 quick transforms
646 Index
direction 227 displaying 227 displaying in front/behind 227 labels 227 line attributes 227 line thickness 227 line type 227 lower specification 227 statistics 227 turning on/off 227 references graph design 27 Marquardt-Levenberg Algorithm 669 regional settings date and time format 77 regression 627, 667 absolute minimum 774 adding equations to graph pages 673, 728, 746 advanced techniques 794 cancelling 690 completion status messages 690 confidence bands 696 constraints, parameter 684, 768 entering equation settings 752 error status messages 720 extending equation curves to axes 673, 727, 746 fit statements 774 generating a regression equation 754 influencing operation 769 iterations 668, 687, 769 local minimum 777 Marquardt-Levenberg algorithm 668 multiple function 791 options 769 parameters 765, 785 prediction bands 696 quitting 693 report 701 results 690 results messages 719 running a regression again 692 saving results 693 scaling x variable 794
solving equations 789 step size 688, 770 tolerance 688, 770, 774 transform functions 762 variables 755 weight variables 763 weighted regression 780 regression equations entering setting 752 iterations 769 regression examples advanced techniques 794 constructor notation 762 dependencies 785 Lorentzian distribution 773 multiple function 791 weighted regression 780 regression examples: solving equations 789 regression functions multiple function 791 regression options entering 769 iterations 769 regression equations 752 regression overview 667 regression results ANOVA table 704 coefficients 703 confidence interval 708 confidence interval for the regression 709 constant variance test 706 constants 703 Cooks Distance test 708 DFFITS 708 diagnostics 707 Durbin-Watson statistic 705 F statistic 704 influence diagnostics 708 leverage 708 normality test 706 P value 704 power 707
647 Index
predicted values 708 prediction interval for the regression 709 PRESS statistic 705 standard error 703 standard error of the estimate 702 statistics 703 sum of squares 704 regression statements bad or missing 720 containing unknown function 720 editing 774 unknown variable 720 Regression Wizard 4, 220, 667, 670 .FIT files 670 about the curve fitter 668 Adding .FIT files to library or notebook 670 cancelling a regression 690 constraints 684 creating new equations 681 Equation Library 671 equation options 683 fit with weight 686 interpreting initial results 690 introduction 667 iterations 687 multiple independent variables 682 parameters 683 running regression from a notebook 680 saving equation changes 682 selecting the equation 671 selecting variables 671 setting graph options 673 setting numeric output options 672 step size 688 tolerance 688 using to add equations to graph pages 673 variable options 682 viewing and editing code 681 viewing initial results 672 watching the fit progress 689 regression: weight variables 760, 781 regression:variables
Poisson distribution 781 regressions fitting data 4 using data in Excel worksheets 102 relational operators transforms 619 removing outliers 92 Report Editor 4 formatting paragraphs 501 formatting text 501 formatting toolbar 501 ruler 498 setting paragraph indents 500 setting tabs 498 report options for nonlinear regression 675, 676 reports creating 495 creating reports 495 creating within Notebook Manager 35 dynamic curve fitting 728 embedding objects 257 exporting 496 global curve fitting 746 inserting date and time 501 linking objects 257 printing 497 regression 701 Report Editor 4 reports 495 setting page size and margins 495 setting ruler units 493 requirements submitting graphs for publication 507 residual tests Durbin-Watson statistic 705 PRESS statistic 705 residuals 676 effect of weighting 764 regression diagnostic results 707 standardized 707 Studentized 707 Studentized deleted 707
648 Index
resize symbols 295 Resizing embedded graphs 178 resizing graphs for publication 510 restoring legends to default settings 296 results completion status messages 719 error status messages 720 linear regressions 222 regression 690 regression messages 690, 719 saving regression 693 viewing constraints 692 retain window states 14 rgbcolor 848 right-clicking custom color 308 hiding graphs 268 selected graphs and objects 240 ROC Curve Analysis 546 rotating 3D graphs 385 3D graphs axes placement 389 angular axis 468 axis titles 440 contour labels 401 labels 289 pie charts 411, 412 text 289 round 849 rows deleting 87 inserting empty 85 selecting 83 sizing 69, 81 titles 88, 92 transposing 87 using as column titles 89, 92 rulers aligning graphs and objects 288
S
sampling data 173 SAS Data Set importing 51 SAS Export File importing 51 satisfying tolerance 719 saving data 33 graphs 33 linear regression results 222 notebook files 33 pages 33 regression equation changes 682 regression results 693 time using the Graph Style Gallery 156 worksheets 33 saving time using the Graph Style Gallery 156 Sax Basic getting Help 519 scalars operators 618 scalars and ranges 618 scale angular 468 axis 3 base/exponent labels 455 category 424, 429 changing 422 common log 423 custom 432 date/time 425 linear 423 logit 424
649 Index
natural log 423 probability 423 probit 423 reciprocal 426 tick labels 458 tick marks 448, 489 types 3, 422, 428, 432 scale type changing 428, 432 scale, axis base/exponent labels 403 contour plots 398 scales Arrhenius 432 Extreme Value Distribution 432 using a date and time scale 431 scaling graphs 277, 279 objects 277, 279 resizing labels/legends automatically with graphs 236 setting aspect ratio option 236 using mouse 278 using Object Properties 279 scatter and line plots examples 374 scatter plots asymmetric error bars 322 creating 311 data format 109 drop lines 208 error bars 317 examples 310, 311 quartile plots 323 range plots 321 symbols 179, 187 scatter plots, 3D creating 377, 378, 379 scientific notation 67 tick labels 455 using as axis values 403, 455 section creating within Notebook Manager 35
editing 37 naming 36, 37 security password protecting notebooks 34 selecting all data in worksheet 83 columns 83 data 83 entire worksheet 83 graph style 113, 150 graph type 109 graphs 163 graphs on page 240 objects on page 240 page objects 240 plots 163 right-clicking graphs 240 rows 83 text on page 240 Selection mode 240 serial numbers 6 set auto recovery options 14 setting axis breaks 438 decimal places 71 equation parameters 213 equations options in the Dynamic Fit Wizard 724 equations options in the Global Fit Wizard 744 equations options in the Regression Wizard 672 line intervals for contour plots 399 macro window options 521 options for nonlinear regression 672, 726, 745 page options 236 passwords 505 passwords for the World Wide Web 505 printing options 103, 105, 238 report options 493 trigonometric units for Quick Transforms 596 setting up graph page format 300 settings 236
650 Index
3D graph view 385 angular axis 468 aspect ratio 236 axis range 427 column statistics 14 error bars 326 frame lines 390 graph defaults 108 object location on page 279 radial axes 472, 475 reference lines 227 regression equations 752 statistics 66 template files 252 worksheet 14 shading 3D graphs 385 between two curves on an area plot 354 shapes lines 190 Shortcut menu hide 268 shortcuts 47 worksheet 48 showing 268 SigmaPlot folders 6 installing 5, 6 registration 6 system requirements 6 using in Windows 9 using OLE2 to edit graphs pasted to other applications 255, 264 SigmaPlot 1.0 and 2.0 (.SPW) importing 51 SigmaPlot Macintosh 4 and 5 Worksheet importing 51 SigmaPlot macros Area Below Curves 533 Batch Process Excel Files 534 Border Plots 535 By Group Data Split 536 Color Transition Values 536
Compute 1st Derivative 537 Frequency Plot 537 Gaussian Cumulative Distribution 538 Insert Graphs into Word 538 Label Symbols 539 Merge Columns 539 Paste to PowerPoint Slide 540 Piper Plots 541 Plotting Polar and Parametric Equations 544 Power Spectral Density 544 Quick Re-Plot 545 Rank and Percentile 545 ROC Curve Analysis 546 Survival Curve 570 Vector Plot 570 SigmaScan (.SPW) importing 51 SigmaStat 1.0 (.SMB) importing 51 SigmaStat transforms 573 simple error bars data format 113 simple error bars and regression data format 113 simple regression data format 113 simple scatter plot data format 113 simple spline curve data format 119, 122 simple straight line data format 122 simple straight line plot data format 119 sin 850 sinh 851 sinp 851 size 852 column statistics 64 symbols 295 values from column for symbols 187 sizing columns and Rows 69
651 Index
columns and rows 81 graphs 277, 279 Notebook Manager 31 objects 277, 279 resizing labels/legends automatically with graphs 236 setting aspect ratio preference 236 using mouse 278 using Object Properties dialog box 279 slices, pie chart edges 272 exploding 411, 412 rotating 411, 412 smoothing unordered XYZ data 604 smoothing data mesh plots 4 snap-to 289 solving differential equations 623 equations 216, 218 equations for x within range 216 functions 216, 218 solving equations 208, 218 sort 852 sorting data 571 source templates for new pages 252 spacing bars 202, 205 bars from different plots 202, 205, 336 special construct functions 805 speed increasing drawing speed 173 SPSS files importing 61 sqrt 628, 853 stacked bar chart 140 stacked bar charts automatic reference lines 226 standard deviation column statistics 64
standard deviation of linear regression coefficients transform 627 standard error column statistics 64 parameter 691 regression results 703 standard error of the estimate regression results 702 Standard percentile method using for box plots 342 standardized residuals regression diagnostic results 707 statements IF function 620 statistical functions 803 statistical summary table results 703 statistical transforms 573 statistics bivariate 622 Durbin-Watson 705 F statistic 704 Options 14 PRESS 705 reference lines 227 setting Options 66 showing/hiding 66 using Excel worksheets 99 worksheet 4 stddev 628, 854 STDDEV function 622 stderr 854 step graph transform 628 step size default value 688 entering 688, 770 Studentized deleted residuals regression results 707 view in regression reports 676 Studentized residuals regression diagnostic results 707 view in regression reports 676 style
652 Index
Graph Style Gallery 158 styles graph 113 using the Graph Style Gallery 156, 157 subblock 855 Submission Assistant 512 editing a profile 513 submitting graphs to journals 507, 510 Subscript 289 suffixes/prefixes contour labels 401 tick labels 458 sum 855 column statistics 64 sum of squares absolute minimum 779 local maximum 779 regression results 704 Superscript 289 survival curves creating 570 switching between date and time and numeric display 82 symbols alphanumeric 183 assigning to worksheet 198 box plots 342 changing edge color 179 changing edge thickness 179 changing fills 179 changing size 179 changing type 179 characters used as 183 color incrementing 181 custom incrementing schemes 198 dot/crosshair color 179 fills 272 Greek 289 increment customizing 198, 205 inserting in legends 289 layering lines in front/behind 190 modifying attributes 179
modifying in plots 179, 187 resize 295 restoring to default settings 296 size values from a worksheet column 187 using characters as 183 using macro to label with 539 Systat European office 25 exporting to 63 system requirements 6 Excel workbooks 6 hardware 6
T
tan 856 tanh 856 tden function 857 tdist function 857 template files settings 252 TEMPLATE.JNT 253 templates 14, 282, 284, 285, 286, 305 adding styles to gallery 157 creating 251, 267 Graph Style Gallery 156, 157, 158 JNT files 253 notebook files 251, 267 options 14 page 251, 267 ternary axes about 478 dragging 480 ternary axis direction modifying 484 ternary data normalizing 596 ternary graphs changing axis direction 484 changing scale direction 484 changing scale type 483 creating 417 data for multiple -curve plot 143
653 Index
data for single plot curve 143 data format 109 defined 143 line plots 417 line/scatter plots 417 modifying plots 419 modifying tick labels 491 modifying tick marks 489 multiple-curve plots 143 normalizing data for 596 plot data set 417 scatter plots 417 selecting data 417 selecting worksheet data 596 third-column data 143 ternary triplets data for 143 testing constant variance 675 normality 675 text adding to page 289 alignment 289 editing contour labels 406 editing tick labels 455 entering 289, 291 formatting 291 grouping/ungrouping 281 labels to page 289 rotating 289 subscript 289 superscript 289 using as plot symbols 183 text files exporting worksheets 63 importing 60 text labels creating 289 text mode entering non-keyboard characters 289 Text Properties dialog box changing text lables 453 thickness
axis lines 434 contour lines 394 grid lines 465 line 190, 273 radial axes 474 reference lines 227 tick marks 451, 489 thickness plot pattern lines 194 Three dimensional 22 tick intervals Date and Time 447 tick labels custom 402, 462 date and time 402 date and time axes 459 editing text 455 engineering notation 455 factors 455 formatting numeric tick labels 455 modifying 458 modifying ternary 491 numeric notation 402 prefix/suffix 458 radial axes 475 scientific notation 455 text attributes 453 time and date 457 using from a worksheet column 402, 462 tick line options 451 tick mark direction 450 tick marks customizing intervals 448 date and time axes 447 hiding 441 intervals assigned to a worksheet 448 labels 458 length 451, 489 modifying 448, 489 modifying ternary 489 turning on/off 451 viewing 441 ticks specifying direction 450
654 Index
time and date see also 457 tinv function 858 tips and tricks avoid repeating the same steps 156 for publication 510 titles axis 442 column 88, 92 column and row titles dialog box 88 graph 166 hiding graph 269 row 88, 92 using cells as column or row titles 92 using worksheet columns as row titles 90 using worksheet rows as column titles 89 tolerance :entering 770 default setting 688 entering 688 reducing 774 satisfying 719 toolbars drawing 270 Excel 101 using Macro Window toolbar 519 Tools menu draw arrow 270 draw box 270 draw ellipse 270 draw line 270 Tools menu commands Text 291 text 289 total 859 TOTAL function 621, 622 Trajectory plots symbols 179 trajectory plots creating 377, 378 symbols 187 transform order of precedence 617
transform components relational operators 616 scalars & ranges 618 transform operators 616 variables 619 transform examples 620 analysis of variance table 621 anova table 621 bivariate statistics 622 coefficient of determination for nonlinear regressions 627 control chart 628 differential equation solving 623 fractional defective control chart 628 F-test to determine statistical improvement in regression 625 linear regression parameters 627 linear regression standard deviations 627 trapezoidal rule beneath a curve 622 transform functions 799 arguments 799 defining variables 762 descriptions 800 multiple regression 791 transform functions and examples 799 transform operators 616 arithmetic 618 defining variables 762 logical 620 order of operation 617 ranges & scalars 618 relational 619 transform variables relational operators 616 transform operators 616 transforms ANOVA.XFM 621 applying to data 4 AREA.XFM 622 arguments 799 as column titles 596 BIVARIAT.XFM 622 defined 4
655 Index
DIFFEQN.XFM 623 duplicate column titles 44 extended 591 F_TEST.XFM 625 from SigmaStat 573 function descriptions 800 functions 799 histogram.xfm 233 normalize ternary data 596 operators 616 R2.XFM 627 ranges & scalars 618 statistical 573 STDV_REG.XFM 627 using data in Excel worksheets 102 using transform language 608 variables 619 transparent mesh creating plots 381 selecting shading 381 Transpose Paste shortcut 48 transposing rows and columns 87 trapezoidal rule 533 trapezoidal rule transform 622 trigonometric functions 801 trigonometric units setting 596 Tukey plot 341 turning on/off axes 433 contour labels 401 contour lines 394 grids 466 insertion mode 50 radial axes 472, 475 radial labels 475 reference lines 227 tick marks 451 two dimensional 15, 21 types graph 109 lines 190
types of graphs 1
U
undo 15 setting for worksheets 44 ungrouping legends 299 objects/text 281 units of measurement page 303 unprotecting Excel workbooks 98 updating object links 264 user accounts 6 user folders 6 user-defined command with bubble plots 145 differential equations 623 F-test 625 user-defined functions creating 526 in macros 525 user-defined transforms 608 function descriptions 800 UserDialog Editor 523 using the equation solver 216, 218
V
values angular 142 axis range 398 axis scale 398 bucket 229 minimum 64 minimum and maximum 64 minimum positive 64 missing 64 radial 142 Z data for contour plots 398 values, axis
656 Index
factors 455 labels 458 numeric labels 403 range 398, 427 time and date 404, 457 variable bar widths 205 variables :dependent 755 defining 761 entering 755 independent 755 relational operators 619 scaling large values 794 unknown 720 weight variable 763, 781 variables: weight variable 760 VBA creating macros using 517 vector formats 509 vector plots 2D 365, 366 creating 570 modifying arrowheads 368 vertical bar chart data format 109 vertical dot plot data format 113 vertical point plot data format 113 view 3D graphs 385 View menu commands 245 drawing toolbar 270 full screen 245 zoom 245 viewing 433 axes 433 axis titles 441 column statistics 4, 64 column statistics in Excel worksheets 99 constraints, parameter 692 contour lines 394 data for embedded graphs 178
full page 245 graphs 245, 268 graphs on page 267, 268 inserted objects as icons 262 linear regression results 222 Macro Window 517 notebook files 38, 39 notebook items 38, 39 object links 264, 266 objects as icons 258 pasted objects as icons 261 tick marks 441 toolbars 101
W
water fall plots creating 379 waterfall plots examples 376 WebViewer 503 publishing graphs on the World Wide Web 503 Weibull scale 432 weight variables 763 entering 755 non-uniform errors 794 norm and residual changes 764 regression 781 when to use 760, 764 weighted regression regression examples 780 weight variables 764 whiskers box plots 341 widths box widths 342 whisker caps 341 Wizards histogram 233 Regression 4 wizards histogram 229 worksheet
657 Index
changing grid color and thickness 70 column statistics 64, 99 column titles 88, 92 column type labels 462 deleting columns and rows 87 entering data 41, 103 Excel 96, 103 exporting as non-notebook file 62, 237 exporting data 103, 104 going to a cell 47 importing data 51, 60, 92 inserting columns and rows 85 inserting graphic cells 198 inserting symbol size values 187 insertion mode 50 moving around 47 moving between notebooks 37, 39 moving data 85 naming 36, 37 opening 39 overwrite mode 50 previewing before printing 104 printing 103, 104 printing column statistics 104 right-click pop-up menu shortcuts 48 row titles 88, 92 selecting all data 83 selecting data 83 selecting entire 83 set data feedback colors 71 setting Options 14 setting printing options 105 shortcuts 47 size 3 sorting data 571 statistics 4 Statistics, showing/hiding 66 transposing rows and columns 87 worksheet functions overview 800 worksheets creating within Notebook Manager 35 entering data 49
exporting as text files 63 formatting date and time data 80 formatting numeric data 80 formatting text 80 macro to split data into multiple columns 536 opening 42 setting Day Zero 77 setting decimal places 71 wrap text 80 World Wide Web publishing graphs to 504 setting passwords to protect your data 505 using the WebViewer 503 wrapping text in worksheet cells 80
X
X,Y values bubble plots 145 X,Y, many Z Contour plots 147 Mesh plots 147 X,Y,Z values contour plots 147 x25 859 x50 860 x75 861 xatymax 862 xwtr 863 XY error bars 329
Z
zooming in/out on graphs 245