Ab initio

Transform components: We have an total of 13 transformation components.

Which are
1. Filter by expression.
2. Reformat.
3. Roll-up.
4. Join.
5. Normalizer.
6. Ded-up sorted.
7. Scan.
8. Multi reformat.
9. Aggregate.
10.Match sorted.
11.Combine.
12.Split.
13.Fuse.

In this all 13 transformation components we are using 9 of them mostly. They are

1. Filter by expression.
2. Reformat.
3. Roll-up.
4. Join.
5. Normalizer.
6. Scan.
7. Multi reformat.
8. Aggregate.
9. Ded-up sorted.
1) Filter by expression: by using this transformation component we can get
the only output data which satisfy the filter condition.

 This will work as same as the filter transformation in informatics.

 This will also works as router transformation in informatics if we are

giving more than one filter condition.

 We can get the different outputs by giving the different filter

conditions.

 We have one output and one deselected port. The data which satisfy
the condition will go to output port and data which doesn’t satisfies
the condition will go to deselect port.

 For every output we have error and reject ports. Which will the details
of error and the records which got rejected.

 We have reject threshold option also.

o Abort on first rejection: it will stops the execution of the

graph on the very first reject event occurs.

o Use ramp/limit: in this option we can give a limit to the

number of errors. If the number of errors exceeds the given
value then the graph should stops the execution. If the number
of errors are within the limit value then the graph should
continue the execution.

o Never abort: If we are using this option then the graph will not
stops the execution if the more than 50% of records also get
rejected.
 Input:
Emp id Ename Dep Manager salary
t
TSIPL1538 JANI ETL SAGAR 20000
P&S
TSIPL1539 ARAVIND ETL SAGAR 25000
P&S
TSIPL1540 KONDA ETL SAGAR 25000
REDDY P&S

Output:
Empid Ename Dept Manager salary
TSIPL1539 ARAVIND ETL SAGAR 25000
P&S
TSIPL1540 KONDA ETL SAGAR 25000
REDDY P&S

Note: filter condition: salary = 25000;

2) Reformat: By using this transformation component we can create some

other ports which are not present in the input. We can calculate the data for
new port by using the existed port data.

 This transformation component works as similar to the expression

transformation in informatica.

 By using this transformation we can add or delete the fields.

 We can increase the number of output ports by increasing the count

parameter.

 We can give get different outputs through the different output ports by
using the DML expressions.
  We can find out the value of the non-existing port by using the
existing port data for the individual record.

 In this reformat we have select parameter which can act as filter by

expression component.

 This select parameter will act as where clause in SQL. The only data
which will satisfy the select parameter condition will come into the
transformation component.

 And we have output index and output indexes also.

 If we are using single condition then we use output index and if we

are using multiple conditions then we will use output indexes option.

 The data which satisfies the output index condition will go through
the transformation component and execute the code which we have
specify in the transform option.

 First the select statement will execute then the index parameter will
execute.

INPUT:

Emp id Ename Dept Manager Salary

TSIPL1538 JANI ETL SAGAR 20000
P&S
TSIPL1539 ARAVIND ETL SAGAR 25000
P&S
TSIPL1540 KONDA ETL SAGAR 25000
REDDY P&S
 OUTPUT:
EMP ID ENAME DEP MANAG SALA ANNU BON
T ER RY AL US
SALAR
Y
TSIPL15 JANI ETL SAGAR 20000 240000 2000
38 P&S
TSIPL15 ARAVI ETL SAGAR 25000 300000 2500
39 ND P&S
TSIPL15 KONDA ETL SAGAR 25000 300000 2500
40 REDDY P&S

Note: annual salary = salary *12; Bonus = salary * 0.1

3) Roll-up: By using this transformation component we can find out the value
for non-existed port by using the data of existed port by grouping the data.

 This transformation component will act as same as aggregate

component in informatics.

 In this transformation we can use the aggregate functions by grouping

the one specified column.

 It will generates the output by grouping the data only.

 For example if we have 1000 records in one table and these 1000
records have 10 different department numbers then in the output we
have only 10 different records if we group the data by department
number.
 INPUT:

Empid Ename Deptno Manager Salary

TSIPL153 JANI 10 SAGAR 20000
8
TSIPL153 ARAVIND 10 SAGAR 25000
9
TSIPL154 KONDA 20 VIDYA 25000
0 REDDY SAGAR

TSIPL154 SHAIK 20 VIDYA 15000

1 SAGAR

OUTPUT:
DEPTNO MANAGE SALARY MIN MAX AVG
R SAL SAL SAL
10 SAGAR 45000 20000 25000 22500
20 VIDYA 50000 15000 25000 20000
SAGAR

4) Join: By using this component we can join the data of two or more than two
inputs by specifying one key.

 This transformation component works as same as joiner component in

informatica.

 By using this component we can join the data from all the inputs
based upon a particular condition.

 We can use all types of sql joins in this join transformation

component.
  We can use inner join, left outer join, right outer join, full outer join
by using the true and false conditions at record match required
parameter.

 We need to specify the column by which we are joining is sorted or

not.

 If the field name was different in different tables then we need specify
it at over-ride key.

 We have an parameter called as driving key. By using this key we can

send the data in table as well as in external.

 By using the driving key non-driving port data will be stored in

external memory and the driving port data will be saved in internal
memory.

 Generally we store the data which have more records at external

memory.

 While joining the tables the number of records should be same.

 We can create and use user defined functions also in this component.

Input-1
Empi Ena Jo Dep sal
d me b tno ary
Tsipl1 Jani Et 10 200
538 l 00
p
&
s
Tsipl1 Arav Et 10 250
539 ind l 00
p
&
s
 Empid 1 Manager Hiredate
Tsipl1538 sagar 4-june
Tsipl1539 Sagar 6-june
Tsipl1540 Sagar 15-march
Tsipl1541 Sagar 4-june
Tsipl1 Kon Et 10 250
540 da l 00
redd p
y &
s
Tsipl1 Shai Et 10 150
541 k l 00
p
&
s

Input-2

Output:
Empid Ename Job Deptno Salary Manager hiredate

Tsipl1538 Jani Etl p&s 10 20000 sagar 4-june

Tsipl1539 Aravind Etl p&s 10 25000 Sagar 6-june

Tsipl1540 Konda Etl p&s 10 25000 Sagar 15-march
reddy
Tsipl1541 Shaik Etl p&s 10 15000 Sagar 4-june

Note: empid = empid 1

5) Scan: By using this transformation component we can get the cumulative

data of the particular specified field.

 This component generates a series of cumulative output for groups of

data records.

 Scan also can be used for some functions like roll-up.

 The main difference between scan and roll-up is scan generates

cumulative reports and roll-up generates summarizes records.

Input:
Empid Ename Job Deptno Mont Salary
h
Tsipl 1538 Jani Etl 10 June 14000
p&s
Tsipl 1539 Aravind Etl 10 June 14000
p&s
Tsipl 1538 Jani Etl 10 July 14000
p&s
Tsipl 1539 Aravind Etl 10 July 14000
p&s
Tsipl01540 Konda Etl 10 July 14000
reddy p&s
Output:

Empid Ename Job Deptno Mont Salary

h
Tsipl 1538 Jani Etl 10 June 14000
p&s
Tsipl 1539 Aravind Etl 10 June 14000
p&s
Tsipl 1538 Jani Etl 10 July 28000
p&s
Tsipl 1539 Aravind Etl 10 July 25000
p&s
Tsipl01540 Konda Etl 10 July 14000
reddy p&s

6) Normalizer: By using this component we can increase or decrease the

number of columns.
 We can create and use the user defined functions in the package
mode.

7) Multi reformat: this component will also works as same as multi reformat.
If we are using more than one output by increasing the count parameter then
we call that reformat as multi reformat.

8) Aggregate: this component will act as same as roll-up but this component
doesn’t have select and output parameter options. We recommend roll-up
 more than aggregate because it provides more operations then the aggregate
component.

9) Dedup sorted: this component is used as a filter for duplicate records. If any
file will have same record for multiple times then by using this component
we can get only one output and the remaining duplicate records are not be
written to the output.