Download the full version and explore a variety of ebooks

or textbooks at https://ebookmass.com

Machine Learning on Geographical Data Using Python

1st Edition Joos Korstanje

_____ Follow the link below to get your download now _____

https://ebookmass.com/product/machine-learning-on-
geographical-data-using-python-1st-edition-joos-korstanje/

Access ebookmass.com now to download high-quality

ebooks or textbooks
We have selected some products that you may be interested in
Click the link to download now or visit ebookmass.com
for more options!.

Machine Learning Guide for Oil and Gas Using Python Hoss
Belyadi

https://ebookmass.com/product/machine-learning-guide-for-oil-and-gas-
using-python-hoss-belyadi/

Machine Learning with Python for Everyone (Addison Wesley

Data & Analytics Series) 1st Edition, (Ebook PDF)

https://ebookmass.com/product/machine-learning-with-python-for-
everyone-addison-wesley-data-analytics-series-1st-edition-ebook-pdf/

Beginner's Guide to Streamlit with Python: Build Web-Based

Data and Machine Learning Applications 1st Edition Sujay
Raghavendra
https://ebookmass.com/product/beginners-guide-to-streamlit-with-
python-build-web-based-data-and-machine-learning-applications-1st-
edition-sujay-raghavendra/

Quantitative Trading Strategies Using Python: Technical

Analysis, Statistical Testing, and Machine Learning
Peng Liu
https://ebookmass.com/product/quantitative-trading-strategies-using-
python-technical-analysis-statistical-testing-and-machine-learning-
peng-liu/
Data-Driven SEO with Python: Solve SEO Challenges with
Data Science Using Python 1st Edition Andreas Voniatis

https://ebookmass.com/product/data-driven-seo-with-python-solve-seo-
challenges-with-data-science-using-python-1st-edition-andreas-
voniatis/

Practical Machine Learning on Databricks (1st Edition)

Debu Sinha

https://ebookmass.com/product/practical-machine-learning-on-
databricks-1st-edition-debu-sinha/

Financial Machina: Machine Learning For Finance: The

Quintessential Compendium for Python Machine Learning For
2024 & Beyond Sampson
https://ebookmass.com/product/financial-machina-machine-learning-for-
finance-the-quintessential-compendium-for-python-machine-learning-
for-2024-beyond-sampson/

Machine Learning for Time Series Forecasting with Python

Francesca Lazzeri

https://ebookmass.com/product/machine-learning-for-time-series-
forecasting-with-python-francesca-lazzeri/

Fundamentals of Machine Learning for Predictive Data

Analytics: Algorithms,

https://ebookmass.com/product/fundamentals-of-machine-learning-for-
predictive-data-analytics-algorithms/
Joos Korstanje

Machine Learning on Geographical Data

Using Python
Introduction into Geodata with Applications and
Use Cases
Joos Korstanje
VIELS MAISONS, France

ISBN 978-1-4842-8286-1 e-ISBN 978-1-4842-8287-8

https://doi.org/10.1007/978-1-4842-8287-8

© Joos Korstanje 2022

Apress Standard

The use of general descriptive names, registered names, trademarks,

service marks, etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the
relevant protective laws and regulations and therefore free for general
use.

The publisher, the authors and the editors are safe to assume that the
advice and information in this book are believed to be true and accurate
at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the
material contained herein or for any errors or omissions that may have
been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Apress imprint is published by the registered company APress

Media, LLC, part of Springer Nature.
The registered company address is: 1 New York Plaza, New York, NY
10004, U.S.A.
Introduction
Spatial data has long been an ignored data type in general data science
and statistics courses. Yet at the same time, there is a field of spatial
analysis which is strongly developed. Due to differences in tools and
approaches, the two fields have long developed in separate
environments.
With the popularity of data in many business environments, the
importance of treating spatial data is also increasing. The goal of the
current book is to bridge the gap between data science and spatial
analysis by covering tools of both worlds and showing how to use tools
from both to answer use cases.
The book starts with a general introduction to geographical data,
including data storage formats, data types, common tools and libraries
in Python, and the like. Strong attention is paid to the specificities of
spatial data, including coordinate systems and more.
The second part of the book covers a number of methods of the field
of spatial analysis. All of this is done in Python. Even though Python is
not the most common tool in spatial analysis, the ecosystem has taken
large steps in user-friendliness and has great interoperability with
machine learning libraries. Python with its rich ecosystem of libraries
will be an important tool for spatial analysis in the near future.
The third part of the book covers multiple machine learning use
cases on spatial data. In this part of the book, you see that tools from
spatial analysis are combined with tools from machine learning and
data science to realize more advanced use cases than would be possible
in many spatial analysis tools. Specific considerations are needed for
applying machine learning to spatial data, due to the specific nature of
coordinates and other specific data formats of spatial data.

Source Code
All source code used in the book can be downloaded from
github.com/apress/machine-learning-geographic-
data-python.
Any source code or other supplementary material referenced by the
author in this book is available to readers on GitHub
(https://github.com/Apress). For more detailed information, please
visit http://www.apress.com/source-code.
Table of Contents
Part I: General Introduction
Chapter 1:​Introduction to Geodata
Reading Guide for This Book
Geodata Definitions
Cartesian Coordinates
Polar Coordinates and Degrees
The Difference with Reality
Geographic Information Systems and Common Tools
What Are Geographic Information Systems
Standard Formats of Geodata
Shapefile
Google KML File
GeoJSON
TIFF/​JPEG/​PNG
CSV/​TXT/​Excel
Overview of Python Tools for Geodata
Key Takeaways
Chapter 2:​Coordinate Systems and Projections
Coordinate Systems
Geographic Coordinate Systems
Projected Coordinate Systems
Local Coordinate Systems
Which Coordinate System to Choose
Playing Around with Some Maps
Example:​Working with Own Data
 Key Takeaways
Chapter 3:​Geodata Data Types
Vector vs.​Raster Data
Dealing with Attributes in Vector and Raster
Points
Definition of a Point
Importing an Example Point Dataset in Python
Some Basic Operations with Points
Lines
Definition of a Line
An Example Line Dataset in Python
Polygons
Definition of a Polygon
An Example Polygon Dataset in Python
Some Simple Operations with Polygons
Rasters/​Grids
Definition of a Grid or Raster
Importing a Raster Dataset in Python
Key Takeaways
Chapter 4:​Creating Maps
Mapping Using Geopandas and Matplotlib
Getting a Dataset into Python
Making a Basic Plot
Plot Title
Plot Legend
Mapping a Point Dataset with Geopandas and Matplotlib
 Concluding on Mapping with Geopandas and Matplotlib
Making a Map with Cartopy
Concluding on Mapping with Cartopy
Making a Map with Plotly
Concluding on Mapping with Plotly
Making a Map with Folium
Concluding on Mapping with Folium
Key Takeaways
Part II: GIS Operations
Chapter 5:​Clipping and Intersecting
What Is Clipping?​
A Schematic Example of Clipping
What Happens in Practice When Clipping?​
Clipping in Python
What Is Intersecting?​
What Happens in Practice When Intersecting?​
Conceptual Examples of Intersecting Geodata
Intersecting in Python
Difference Between Clipping and Intersecting
Key Takeaways
Chapter 6:​Buffers
What Are Buffers?​
A Schematic Example of Buffering
What Happens in Practice When Buffering?​
Creating Buffers in Python
Creating Buffers Around Points in Python
 Creating Buffers Around Lines in Python
Creating Buffers Around Polygons in Python
Combining Buffers and Set Operations
Key Takeaways
Chapter 7:​Merge and Dissolve
The Merge Operation
What Is a Merge?​
A Schematic Example of Merging
Merging in Python
Row-Wise Merging in Python
Attribute Join in Python
Spatial Join in Python
The Dissolve Operation
What Is the Dissolve Operation?​
Schematic Overview of the Dissolve Operation
The Dissolve Operation in Python
Key Takeaways
Chapter 8:​Erase
The Erase Operation
Schematic Overview of Spatially Erasing Points
Schematic Overview of Spatially Erasing Lines
Schematic Overview of Spatially Erasing Polygons
Erase vs.​Other Operations
Erase vs.​Deleting a Feature
Erase vs.​Clip
Erase vs.​Overlay
 Erasing in Python
Erasing Portugal from Iberia to Obtain Spain
Erasing Points in Portugal from the Dataset
Cutting Lines to Be Only in Spain
Key Takeaways
Part III: Machine Learning and Mathematics
Chapter 9:​Interpolation
What Is Interpolation?​
Different Types of Interpolation
Linear Interpolation
Polynomial Interpolation
Nearest Neighbor Interpolation
From One-Dimensional to Spatial Interpolation
Spatial Interpolation in Python
Linear Interpolation Using Scipy Interp2d
Kriging
Linear Ordinary Kriging
Gaussian Ordinary Kriging
Exponential Ordinary Kriging
Conclusion on Interpolation Methods
Key Takeaways
Chapter 10:​Classification
Quick Intro to Machine Learning
Quick Intro to Classification
Spatial Classification Use Case
Feature Engineering with Additional Data
 Importing and Inspecting the Data
Spatial Operations for Feature Engineering
Reorganizing and Standardizing the Data
Modeling
Model Benchmarking
Key Takeaways
Chapter 11:​Regression
Introduction to Regression
Spatial Regression Use Case
Importing and Preparing Data
Iteration 1 of Data Exploration
Iteration 1 of the Model
Iteration 2 of Data Exploration
Iteration 2 of the Model
Iteration 3 of the Model
Iteration 4 of the Model
Interpretation of Iteration 4 Model
Key Takeaways
Chapter 12:​Clustering
Introduction to Unsupervised Modeling
Introduction to Clustering
Different Clustering Models
Spatial Clustering Use Case
Importing and Inspecting the Data
Cluster Model for One Person
Tuning the Clustering Model
 Applying the Model to All Data
Key Takeaways
Chapter 13:​Conclusion
What You Should Remember from This Book
Recap of Chapter 1 – Introduction to Geodata
Recap of Chapter 2 – Coordinate Systems and Projections
Recap of Chapter 3 – Geodata Data Types
Recap of Chapter 4 – Creating Maps
Recap of Chapter 5 – Clipping and Intersecting
Recap of Chapter 6 – Buffers
Recap of Chapter 7 – Merge and Dissolve
Recap of Chapter 8 – Erase
Recap of Chapter 9 – Interpolation
Recap of Chapter 10 – Classification
Recap of Chapter 11 – Regression
Recap of Chapter 12 – Clustering
Further Learning Path
Going into Specialized GIS
Specializing in Machine Learning
Remote Sensing and Image Treatment
Other Specialties
Key Takeaways
Index
About the Author
Joos Korstanje
is a data scientist, with over five years of
industry experience in developing
machine learning tools. He has a double
MSc in Applied Data Science and in
Environmental Science and has extensive
experience working with geodata use
cases. He has worked at a number of
large companies in the Netherlands and
France, developing machine learning for
a variety of tools. His experience in
writing and teaching has motivated him
to write this book on machine learning
for geodata with Python.
About the Technical Reviewer
Xiaochi Liu
is a PhD researcher and data scientist at
Macquarie University, specializing in
machine learning, explainable artificial
intelligence, spatial analysis, and their
novel application in environmental and
public health. He is a programming
enthusiast using Python and R to
conduct end-to-end data analysis. His
current research applies cutting-edge AI
technologies to untangle the causal
nexus between trace metal
contamination and human health to
develop evidence-based intervention
strategies for mitigating environmental exposure.
Part I
General Introduction
© The Author(s), under exclusive license to APress Media, LLC, part of Springer
Nature 2022
J. Korstanje, Machine Learning on Geographical Data Using Python
https://doi.org/10.1007/978-1-4842-8287-8_1

1. Introduction to Geodata
Joos Korstanje1
(1) VIELS MAISONS, France

Mapmaking and analysis of the geographical environment around us have

been present in nature and human society for a long time. Human maps
are well known to all of us: they are a great way to share information
about our environment with others.
Yet communicating geographical instructions is not invented only by
the human species. Bees, for example, are well known to communicate on
food sources with their fellow hive mates. Bees do not make maps, but,
just like us, they use a clearly defined communication system.
As geodata is the topic of this book, I find it interesting to share this
out-of-the-box geodata system used by honeybees. Geodata in the bee
world has two components: distance and direction.
Honeybee distance metrics
– The round dance: A food source is present less than 50 meters from the
hive.
– The sickle dance: Food sources are present between 50 and 150 meters
from the hive.
– The waggle (a.k.a. wag-tail) dance: Food sources are over 150 meters
from the hive. In addition, the duration of the waggle dance is an
indicator of how far over 150 meters the source is located.
Honeybee direction metrics
– Although more complicated, the angle of the dance is known to be an
indicator of the angle relative to the sun that bees must follow to get to
their food source.
 – As the sun changes location throughout the day, bees will update each
other by adapting their communication dances accordingly.
The human counterpart of geographical communication works a bit
better, as we have compasses that point to the magnetic north. Those of
you who are familiar with compass use, for example, on boats, may know
that even using a compass is not a perfect solution.
The magnetic north changes much less than the position of the sun.
What is interesting though is that the magnetic north and the true north
are not located at the same exact place. The true north is a fixed location
on the globe (the so-called North Pole), but compasses are based on
magnetism and therefore point to the magnetic north: a location that
moves a little bit every year.
If you are navigating a ship with a compass, you will constantly need
to do calculations that convert your magnetic direction measurements
into true direction measurements by adding magnetic variation, which is
a value that changes depending on where you are on earth.

Reading Guide for This Book

As you will understand from these two examples, working with geodata is
a challenge. While identifying locations of points by coordinates may
appear simple, the devil really is in the details.
The goal of this book is to go over all those details while working on
example code projects in Python. This should give you the fundamental
knowledge needed to start working in the interesting domain of geodata
while avoiding mistakes. You will then discover numerous ways to
represent geodata and learn to work with tools that make working with
geodata easier.
After laying the basis, the book will become more and more advanced
by focusing on machine learning techniques for the geodata domain. As
you may expect, the specificities of the use of geodata make that a lot of
standards techniques are not applicable at all, or in other cases, they may
need specific adaptations and configurations.

Geodata Definitions
To get started, I want to cover the basics of coordinate systems in the
simplest mathematic situation: the Euclidean space. Although the world
does not respect the hypothesis made by Euclidean geometry, it is a great
entry into the deeper understanding of coordinate systems.
A two-dimensional Euclidean space is often depicted as shown in
Figure 1-1.

Figure 1-1 A two-dimensional Euclidean space. Image by author

Cartesian Coordinates
To locate points in the Euclidean space, we can use the Cartesian
coordinate system. This coordinate system specifies each point uniquely
by a pair of numerical coordinates. For example, look at the coordinate
system in Figure 1-2, in which two points are located: a square and a
triangle.
The square is located at x = 2 and y = 1 (horizontal axis). The triangle
is located at x = -2 and y = -1.
Figure 1-2 Two points in a coordinate system. Image by author
The point where the x and y axes meet is called the Origin, and
distances are measured from there. Cartesian coordinates are among the
most well-known coordinate system and work easily and intuitively in the
Euclidean space.

Polar Coordinates and Degrees

A commonly used alternative to Cartesian coordinates is the polar
coordinate system. In the polar system, one starts by defining one point as
the pole. From this pole starts the polar axis. The graphic in Figure 1-3
shows the idea.
Figure 1-3 The polar system. Image by author
In this schematic drawing, the star is designated as the pole, and the
thick black line to the right is chosen as the polar axis. This system is
quite different from the Cartesian system but still allows us to identify the
exact same points: just in a different way.
The points are identified by two components: an angle with respect to
the polar axis and a distance. The square that used to be referred to as
Cartesian coordinate (2,1) can be referred to by an angle from the polar
axis and a distance.
This is shown in Figure 1-4.
Figure 1-4 A point in the polar coordinate system. Image by author
At this point, you can measure the distance and the angle and obtain
the coordinate in the polar system. Judged by the eye alone, we could say
that the angle is probably more or less 30° and the distance is slightly
above 2. We would need to have more precise measurement tools and a
more precise drawing for more precision.
There are trigonometric computations that we can use to convert
between polar and Cartesian coordinates. The first set of formulas allows
you to go from polar to Cartesian:

The letter r signifies the distance and the letter φ is the angle. You can go
the other way as well, using the following formulas:

As a last part to cover about degrees, I want to mention the equivalence

between measuring angles in degrees and in radians. The radian system
may seem scary if you are not used to it, but just remember that for every
possible angle that you can measure (from 0 to 360) there is a
corresponding notation in the radian system. Figure 1-5 shows this.
Figure 1-5 Radians vs. degrees. Image by author

The Difference with Reality

In reality, you never work with Euclidean space on a map. This is because
the world is not flat, but rather a sort of sphere. First of all, it needs to be
considered that the object is in three dimensions. More importantly,
distances from one point to another need to take into account the specific
curvature of the earth at that point. After all, to make it even more
difficult, the earth unfortunately is not a perfectly round ball.
In the real world, things have to be much more complicated than in
the Euclidean examples. This is done by the Geographic Coordinate
System which is adapted to our ellipsoidal reality. In this system, we
usually measure a point by a combination of latitude and longitude.
Latitude indicates how high or low on the globe you are with respect
to the equator. Longitude tells us how much left or right on the globe you
are with respect to the Greenwich meridian.
The earth is split into four quadrants from the zero point at the
intersection of the equator and the Greenwich meridian. You have north
and south and east and west together making up four quadrants. The
North Pole has a latitude of 90 degrees North, and the South Pole is 90
degrees South. Longitude ranges from 180 degrees West to 180 degrees
East.
 Degrees do not have decimals, but rather can be cut up into minutes.
One degree consists of 60 minutes, and one minute consists of 60
seconds.

Geographic Information Systems and Common Tools

As you must understand by now, geodata is an easy way into a headache if
you do not have tools that do all the conversion work for you. And we are
lucky, as many such tools exist. In this part, let’s have a look at a few of the
most commonly used tools together with some advantages and
disadvantages of each of them.

What Are Geographic Information Systems

GIS or Geographic Information Systems are a special type of database
system that is made specifically for geographic data, also called geodata.
Those database systems are developed in such a way that problems like
coordinate systems and more are not a problem to be solved by the user.
It is all done inherently by the system. GIS also stands for the industry
that deals with those information systems.

ArcGIS
ArcGIS, made by ESRI, is arguably the most famous software package for
working with Geographic Information Systems. It has a very large number
of functionalities that can be accessed through a user-friendly click-
button system, but visual programming of geodata processing pipelines is
also allowed. Python integration is even possible for those who have
specific tasks for which there are no preexisting tools in ArcGIS. Among
its tools are also AI and data science options.
ArcGIS is a great software for working with geodata. Yet there is one
big disadvantage, and that is that it is a paid, proprietary software. It is
therefore accessible only to companies or individuals that have no
difficulty paying the considerably high price. Even though it may be worth
its price, you’ll need to be able to pay or convince your company to pay for
such software. Unfortunately, this is often not the case.

QGIS and Other Open Source ArcGIS Alternatives

Open source developers have jumped into this open niche of GIS systems
by developing open source (and therefore free to use) alternatives. These
include QGIS, GRASS GIS, PostGIS, and more.
The clear advantage of this is that they are free to use. Yet their
functionality is often much more limited. In most of them, users have the
ability to code their own modules in case some of the needed tools are not
available.
This approach can be a good fit for your need if you are not afraid to
commit to a system like QGIS and fill the gaps that you may eventually
encounter.

Python/R Programming
Finally, you can use Python or R programming for working with geodata
as well. Programming, especially in Python or R, is a very common skill
among data professionals nowadays.
As programming skills were less well spread a few years back, the
boom in data science, machine learning, and artificial intelligence has
made languages like Python become very commonly spread throughout
the workforce.
Now that many are able to code or have access to courses to learn how
to code, the need for full software becomes less. The availability of a
number of well-functioning geodata packages is enough for many to get
started.
Python or R programming is a great tool for treating geodata with
common or more modern methods. By using these programming
languages, you can easily apply tools from other libraries to your geodata,
without having to convert this to QGIS modules, for example.
The only problem that is not very well solved by programming
languages is long-term geodata storage. For this, you will need a database.
Cloud-based databases are nowadays relatively easy to arrange and
manage, and this problem is therefore relatively easily solved.

Standard Formats of Geodata

As you have understood, there are different tools and programming
languages that can easily deal with geodata. While doing geodata in
Python in this book, we will be generally interested much more in data
processing than in data storage.
 Yet a full solution for geodata treatment cannot rely on treatment
alone. We also need a data format. You will now see a number of common
data formats that are used very widely for storing geographical data.

Shapefile
The shapefile is a very commonly used file format for geodata because it
is the standard format for ArcGIS. The shapefile is not very friendly for
being used outside of ArcGIS, but due to the popularity of ArcGIS, you will
likely encounter shapefiles at some point.
The shapefile is not really a single file. It is actually a collection of files
that are stored together in one and the same directory, all having the
same name. You have the following files that make up a shapefile:
– myfile.shp: The main file, also called the shapefile (confusing but true)
– myfile.shx: The shapefile index file
– myfile.dbf: The shapefile data file that stores attribute data
– myfile.prj: Optional file that stores spatial reference and projection
metadata
As an example, let’s look at an open data dataset containing the
municipalities of the Paris region that is provided by the French
government. This dataset is freely available at
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f
2c318a0971e8faedb7b5675d6
On this website, you can download the data in SHP/L93 format, and
this will allow you to download a directory with a zip file. Figure 1-6
shows what this contains.

Figure 1-6 The inside of the shapefile. Image by author Data source: Ministry of
DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
 As you can see, there are the .shp file (the main file), the .shx file (the
index file), the .dbf file containing the attributes, and finally the optional
.prj file.
For this exercise, if you want to follow along, you can use your local
environment or a Google Colab Notebook at
https://colab.research.google.com/.
You have to make sure that in your environment, you install
geopandas:

!pip install geopandas

Then, make sure that in your environment you have a directory called
Communes_MGP.shp in which you have the four files:
– Communes_MGP.shp
– Communes_MGP.dbf
– Communes_MGP.prj
– Communes_MGP.shx
In a local environment, you need to put the “sample_data” file in the
same directory as the notebook, but when you are working on Colab, you
will need to upload the whole folder to your working environment, by
clicking the folder icon and then dragging and dropping the whole folder
onto there. You can then execute the Python code in Code Block 1-1 to
have a peek inside the data.

import geopandas as gpd

shapefile =
gpd.read_file("sample_data/Communes_MGP.shp")
print(shapefile)
Code Block 1-1 Importing the shapefile

You’ll see the result in Figure 1-7.

Figure 1-7 The data in Python. Image by author Data source: Ministry of DINSIC.
Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
To make something more visual, you can use the code in Code Block 1-
2.

shapefile.plot()
Code Block 1-2 Plotting the shapefile

You will obtain the map corresponding to this dataset as in Figure 1-8.
Figure 1-8 The map resulting from Code Block 1-2. Image by author Data source:
Ministry of DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)

Google KML File

You are probably familiar with Google Earth: one of the dominating map-
based applications of our time. Google has popularized the KML file for
geodata. It is an XML-based text file that can contain geometry data.
The .KMZ file is a compressed version of the KML file. You can
decompress it using any unzipping tool and obtain a KML file.
As an example, let’s look at the exact same database as before, which
is located at France’s open geodata platform:
Ministry of DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f
2c318a0971e8faedb7b5675d6, updated on 1 July 2016. Open Licence
2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
 In the resources part, you’ll see that this map of the Paris region’s
municipalities is also available in the KML format. Download it and you’ll
obtain a file called Communes_MGP.kml.
If you try opening the file with a text editor, you’ll find that it is an XML
file (very summarized, XML is a data storage pattern that can be
recognized by many < and > signs).
Compared to the shapefile, you can see that KML is much easier to
understand and to parse. A part of the file contents is shown in Figure 1-9.
Figure 1-9 The KML file content. Image by author Data source: Ministry of DINSIC.
Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
To get a KML file into Python, we can again use geopandas. This time,
however, it is a bit less straightforward. You’ll also need the Fiona package
to obtain a KML driver. The total code is shown in Code Block 1-3.

import fiona
gpd.io.file.fiona.drvsupport.supported_drivers['KML']
= 'rw'

import geopandas as gpd

kmlfile = gpd.read_file("Communes_MGP.kml")
print(kmlfile)
Code Block 1-3 Importing the KML file

You’ll then see the exact same geodataframe as before, which is shown
in Figure 1-10.

Figure 1-10 The KML data shown in Python. Image by author Data source: Ministry of
DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
 As before, you can plot this geodataframe to obtain a basic map
containing the municipalities of the area of Paris and around. This is done
in Code Block 1-4.

kmlfile.plot()
Code Block 1-4 Plotting the KML file data

The result is shown in Figure 1-11.

Figure 1-11 The plot resulting from Code Block 1-4. Screenshot by author Data source:
Ministry of DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)

An interesting point here is that the coordinates do not correspond

with the map that was generated from the shapefile. If you’ve read the
first part of this chapter, you may have a hinge on how this is caused by
coordinate systems. We’ll get into this in much more detail in Chapter 2.

GeoJSON
The json format is a data format that is well known and loved by
developers. Json is much used in communication between different
information systems, for example, in website and Internet
communication.
The json format is loved because it is very easy to parse, and this
makes it a perfect storage for open source and other developer-oriented
tools.
Json is a key-value dataset, which is much like the dictionary in
Python. The whole is surrounded by accolades. As an example, I could
write myself as a json object as in this example:

{ 'first_name': 'joos',
'last_name': 'korstanje',
'job': 'data scientist' }

As you can see, this is a very flexible format, and it is very easy to
adapt to all kinds of circumstances. You might easily add GPS coordinates
like this:

{ 'first_name': 'joos',
'last_name': 'korstanje',
'job': 'data scientist',
'latitude': '48.8566° N',
'longitude': '2.3522° E' }

GeoJSON is a json-based format that defines a specific, standardized

way to deal with storing coordinates (not just points but also lines and
polygons) in the json format.
The Paris municipalities map that you’ve downloaded before is also
available in the geojson format. Download it over here (click GeoJSON in
the Resources part):
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b
94f2c318a0971e8faedb7b5675d6
You will obtain a file called Communes_MGP.json. When opening it
with notepad or any other text editor, you’ll see that it is a json format
(shown in Figure 1-12). Of course, it is the exact same data: only the
storage format changes.
Figure 1-12 The content in json format. Image by author Data source: Ministry of
DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
You can get a GeoJSON file easily into the geopandas library using the
code in Code Block 1-5.

import geopandas as gpd

geojsonfile = gpd.read_file("Communes_MGP.json")
print(geojsonfile)
Code Block 1-5 Importing the geojson

As expected, the data looks exactly like before (Figure 1-13). This is
because it is transformed into a geodataframe, and therefore the original
representation as json is not maintained anymore.
Figure 1-13 The geojson content in Python. Image by author Data source: Ministry of
DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
You can make the plot of this geodataframe to obtain a map, using the
code in Code Block 1-6.

geojsonfile.plot()
Code Block 1-6 Plotting the geojson data

The resulting plot is shown in Figure 1-14.

Figure 1-14 The plot resulting from Code Block 1-6. Image by author Data source:
Ministry of DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/8fadd7040c4b94f2c318a0971e8faedb7b5675d6,
updated on 1 July 2016. Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)

TIFF/JPEG/PNG
Image file types can also be used to store geodata. After all, many maps
are 2D images that lend themselves well to be stored as an image. Some of
the standard formats to store images are TIFF, JPEG, and PNG.
– The TIFF format is an uncompressed image. A georeferenced TIFF
image is called a GeoTIFF, and it consists of a directory with a TIFF file
and a tfw (world file).
– The better-known JPEG file type stores compressed image data. When
storing a JPEG in the same folder as a JPW (world file), it becomes a
GeoJPEG.
– The PNG format is another well-known image file format. You can make
this file into a GeoJPEG as well when using it together with a PWG
(world file).
 Image file types are generally used to store raster data. For now,
consider that raster data is image-like (one value per pixel), whereas
vector data contains objects like lines, points, and polygons. We’ll get to
the differences between raster and vector data in a next chapter.
On the following website, you can download a GeoTIFF file that
contains an interpolated terrain model of Kerbernez in France:
https://geo.data.gouv.fr/en/datasets/b0a420b9e003
d45aaf0670446f0d600df14430cb
You can use the code in Code Block 1-7 to read and show the raster file
in Python.

pip install rasterio

import rasterio
from rasterio.plot import show
fp = r'ore-kbz-mnt-litto3d-5m.tif'
img = rasterio.open(fp)
show(img)
Code Block 1-7 Read and show the raster data

Note Depending on your OS, you may obtain a .tiff file format rather
than a .tif when downloading the data. In this case, you can simply
change the path to become .tiff, and the result should be the same. In
both cases, you will obtain the image shown in Figure 1-15.
Figure 1-15 The plot resulting from Code Block 1-7. Image by author Data source:
Ministry of DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/b0a420b9e003d45aaf0670446f0d600df14430cb,
updated on “unknown.” Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)

It is interesting to look at the coordinates and observe that this file’s

coordinate values are relatively close to the first file.

CSV/TXT/Excel
The same file as used in the first three examples is also available in CSV.
When downloading it and opening it with a text viewer, you will observe
something like Figure 1-16.
Figure 1-16 The contents of the CSV file. Image by author Data source: Ministry of
DINSIC. Original data downloaded from
https://geo.data.gouv.fr/en/datasets/b0a420b9e003d45aaf0670446f0d600df14430cb,
updated on “unknown.” Open Licence 2.0 (www.etalab.gouv.fr/wp-
content/uploads/2018/11/open-licence.pdf)
The important thing to take away from this part of the chapter is that
geodata is “just data,” but with geographic references. These can be stored
in different formats or in different coordinate systems to make things
complicated, but in the end you must simply make sure that you have
some sort of understanding of what you have in your data.
You can use many different tools for working with geodata. The goal of
those tools is generally to make your life easier. As a last step for this
introduction, let’s have a short introduction to the different Python tools
that you may encounter on your geodata journey.
Overview of Python Tools for Geodata
Here is a list of Python packages that you may want to look into on your
journey into geodata with Python:

Geopandas
General GIS tool with a pandas-like code syntax that makes it very
accessible for the data science world.

Fiona
Reading and writing geospatial data.

Rasterio
Python package for reading and writing raster data.

GDAL/OGR
A Python package that can be used for translating between different GIS
file formats.

RSGISLIB
A package containing remote sensing tools together with raster
processing and analysis.

PyProj
A package that can transform coordinates with multiple geographic
reference systems.

Geopy
Find postal addresses using coordinates or the inverse.

Shapely
Manipulation of planar geometric objects.

PySAL
Spatial analysis package in Python.

Scipy.spatial
Spatial algorithms based on the famous scipy package for data science.

Cartopy
Package for drawing maps.

GeoViews
Package for interactive maps.
A small reminder: As Python is an open source environment and those
libraries are mainly developed and maintained by unpaid open source
developers, there is always that chance that something changes or
becomes unavailable. This is the risk of working with open source. In
most cases, there are no such big problems, but they can and do
sometimes happen.

Key Takeaways
1. Cartesian coordinates and polar coordinates are two alternative
coordinate systems that can indicate points in a two-dimensional
Euclidean space.

2. The world is an ellipsoid, which makes the two-dimensional

Euclidean space a bad representation. Other coordinate systems exist
for this real-world scenario.

3. Geodata is data that contains geospatial references. Geodata can come

in many different shapes and sizes. As long as you have software
implementation (or the skills to build it), you will be able to convert
between data formats.

4. A number of Python packages exist that do a lot of the heavy lifting

for you.

5. The advantage of using Python is that you can have a lot of autonomy
on your geodata treatment and that you can benefit from the large
number of geodata and other data science and AI packages in the
ecosystem.

6. A potential disadvantage of Python is that the software is open

source, meaning that you have no guarantee that your preferred
libraries still exist in the future. Python is also not suitable for long-
term data storage and needs to be complemented with such a data
storage solution (e.g., databases or file storage).
© The Author(s), under exclusive license to APress Media, LLC, part of Springer
Nature 2022
J. Korstanje, Machine Learning on Geographical Data Using Python
https://doi.org/10.1007/978-1-4842-8287-8_2

2. Coordinate Systems and Projections

Joos Korstanje1
(1) VIELS MAISONS, France

In the previous chapter, you have seen an introduction to coordinate

systems. You saw an example of how you can use Cartesian coordinates as
well as polar coordinates to identify points on a flat, two-dimensional
Euclidean space. It was already mentioned at that point that the real-
world scenario is much more complex.
When you are making maps, you are showing things (objects, images,
etc.) that are located on earth. Earth does not respect the rules that were
shown in the Euclidean example because Earth is an ellipsoid: a ball form
that is not perfectly round. This makes map and coordinate system
calculations much more complex than what high-school mathematics
teaches us about coordinates.
To make the problem clearer, let’s look at an example of airplane
navigation. Airplane flights are a great example to illustrate the problem,
as they generally cover long distances. Taking into account the curvature
of the earth really doesn’t matter much when measuring the size of your
terrace, but it does make a big impact when moving across continents.
Imagine you are flying from Paris to New York using this basic sketch
of the world’s geography. You are probably well aware of such an
organization of the world’s map on a two-dimensional image.
A logical first impression would be that to go from Madrid to New
York in the quickest way, we should follow a line parallel from the latitude
lines. Yet (maybe surprisingly at first) this is not the shortest path. An
airplane would better curve via the north!
The reason for this is that the more you move to the north, the shorter
the latitude lines actually are. Latitude lines go around the earth, so at the
North Pole you have a length of zero, and at the equator, the middle line is
the longest possible. The closer to the poles, the shorter the distance to go
around the earth.
As this example takes place in the northern hemisphere, the closest
pole is the North Pole. By curving north on the northern hemisphere
(toward the pole), an airplane can get to its destination with fewer
kilometers. Figure 2-1 illustrates this.

Figure 2-1 Airplane routes are not straight on a mapImage adapted from
https://en.wikipedia.org/wiki/World_map#/media/File:Blue_Marble_2002.png (Public
Domain Image. 10 February 2002)

Let’s now consider an example where you are holding a round soccer
ball. When going from one point to another on a ball, you will intuitively
be able to say which path is the fastest. If you are looking straight at the
ball, when following your finger going from one point to another, you will
see your hand making a shape like in Figure 2-2.
Figure 2-2 The shortest path on a ball is not a straight line in two-dimensional view.
Image by author
When making maps, we cannot plot in three dimensions, and we,
therefore, need to find some way or another to put a three-dimensional
path onto a two-dimensional image.
Many map makers have proposed all sorts of ways to solve this
unsolvable problem, and the goal of this chapter is to help you understand
how to deal effectively with those 3D to 2D mapping distortions that will
be continuously looking to complexify your work on geodata.

Coordinate Systems
While the former discussion was merely intuitive, it is now time to slowly
get to more official definitions of the concepts that you have seen. As we
are ignoring the height of a point (e.g., with respect to sea level) for the
moment, we can identify three types of coordinate systems:
– Geographic Coordinate Systems
– Projected Coordinate Systems
– Local Coordinate Systems
Let’s go over all three of them.

Geographic Coordinate Systems

Other documents randomly have
different content
traveller and writer, who translated them into his national tongue. He
was much delighted with the information which I gave him, and told
me that my translation enabled him to interpret something of the
spirit of the great Tibetan epic.
We stayed one night at that temple, and on the following day
proceeded along the banks of the river Ham-hung-gi-chu (shoe-
dropping river) and reached a place which contained a temple called
Gyang-tak-gonpa. This temple is dedicated to Dorje Karmo, the
Goddess named White Vajra. The place is situated about one mile off
the road and near by is a postal station named Darchen Tazam. This
station contains about thirty houses built of stone, besides about a
dozen tents pitched here and there. It is a business as well as a
revenue centre for the whole district. I lodged at one of the houses,
and here the guide took leave of me. That night I performed my
usual religious meditation, and on the morning of the following day
my pilgrim companions rejoined me.
The station lies on a steppe between the north-western corner of
Lake Mānasarovara and the north-eastern corner of Lake Lakgal. On
the following day our party left the station, and proceeded in a
south-easterly direction, to the west of Mānasarovara. We advanced
in the same direction the next day, till we reached the foot of a
snowy peak named Bon-Ri. This is, as I have mentioned before, a
place sacred to the Bon, or ancient religion of Tibet. I saw a big
temple in the place, which I found to be not a temple belonging to
that old religion, as I had expected, but one belonging to the New
Sect. It looked a magnificent establishment as seen from a distance,
but we did not go near it. This neighborhood produced various kinds
of mushrooms, and some which were growing in damp places were
gathered by the women of the party. They collected large quantities
of the fungus, which was fried with butter and eaten with salt. I
tasted it and found it delicious. By that time we had left the limits of
the sacred region, and my male companions no longer considered
themselves as pilgrims, but as men who had to face the stern
realities of the material world. They declared that they must resume
their worldly business, and proposed to start by shooting deer. It
seemed to me that their shooting not infrequently included
extraordinary kinds of game, and I suspected, on good grounds, that
the three brothers had now and then turned highwaymen and either
robbed or murdered travellers. I therefore began to be afraid of
them, and thought that I had better separate myself from them on
some plausible pretext, and without awakening their suspicion.
On the following day we reached the brow of a hill, and there one of
the brothers in my presence shot an animal called in Tibet changku.
The shooting was done merely for pleasure and not with the object
of eating its flesh or using its skin. The changku, or wolf, resembles
a large species of dog with rather thin fur, which in summer turns a
fine brownish color. In winter the color is said to be a whitish grey.
The ears are erect and the face appears ferocious. It is said that this
wild animal will attack solitary travellers and even kill them. When
the brothers brought down the animal their eyes gleamed with
delight, and I secretly thought that their eyes would show that same
cruel gleam when they murdered a wealthy traveller.
 CHAPTER XXXI.
An Ominous Outlook.
The next day, September 14th, snow again fell, and so we had to
stay in the same place. The hunting-dogs went out of their own
accord on a rabbit-hunting expedition, and came back with their
mouths stained with blood. They must have hunted down some
rabbits and made a meal of them. The snow ceased, and we left the
place on the following day. Proceeding eastwards, we now came to a
long undulating hill, and soon reached its summit. Here the head of
the family said that our pilgrimage must end at this spot, and when
asked why at this particular place, he pointed to Lake Mānasarovara,
situated to the west, and also to the snow-capped peak of Manri
that stood due south from the middle of the Lake, and told me that
we should here bid farewell and express our good wishes to the
sacred region, for this was the last point where we could have a full
view of the Holy Place, and that we should express in our prayers an
earnest desire to visit this sacred region again in the future. Saying
this, he bowed down and I and all the rest followed his example.
When I thought that I (the first Japanese who had ever come to visit
this district from a remote country thousands of miles away) was
now about to take leave of Lake Mānasarovara after having been in
its neighborhood for several days, a peculiar sensation came over
me, and I stood gazing at the lake for some time. As we were going
down the hill, my host told me that as they had already departed
from the Holy Place they should now earnestly engage themselves in
their worldly pursuits; therefore they thought it time that I should
leave them. We soon reached a little encampment of some twelve or
thirteen tents, and thither I wended my way to observe the
condition of the small community.
Mendicancy was well suited for satisfying my curiosity, and as a
mendicant I entered the encampment. My companions remained in
the same place that day and the next, the brothers occupied in
shooting. On the latter day I was reading a Chinese Buḍḍhist Text,
and the two women were outside engaged in some earnest talk. At
first I did not pay any attention to what they were saying, but when
my ears caught the word ‘Lama’ pronounced several times my
curiosity was awakened. Dawa was saying that she had heard the
Lama, that is myself, say that her mother was probably dead. She
wished, she continued, to ascertain this of the Lama, and so she had
been pressing him for some definite information. Her aunt received
this remark with a laugh. He must have seen, she said, that Dawa
was in love with him, and had therefore told her this fib in joke. She
must not mind what the Lama told her. However, the aunt continued,
her husband had been telling her that he must make the Lama
marry Dawa, and that should he refuse, her husband would kill him.
It was evident that this last portion of the conversation was intended
for my ears, for the aunt spoke in a loud voice.
When I heard that intimidatory warning I at first felt alarmed, but
the next moment I recovered my tranquillity. I thought that if I
should suffer death for having resisted a temptation, my death
would be highly approved by the holy Founder. He would be
displeased if I should disobey my conscience for the mere fear of
death. Internally praying for strength of mind to resist the
temptation, even at the risk of my life, I resumed my reading.
However nothing occurred to me that day, nor the next, when we
struck our tent and proceeded for about five miles close to the brow
of a hill, from which I saw at a short distance what appeared to be
houses, and I was told that this was another postal station called
Tokchen Tazam. Again I visited the place in the disguise of a
mendicant priest. I soon returned and found Dawa alone in the tent;
the rest were all gone out hunting, so she told me. I at once saw
that the conspiracy was developing, and that matters were growing
quite critical.
I concluded that I must do my best to dissuade the girl from
pursuing the object of her misplaced affection. Some spiritual affinity
must have brought me into the company of this girl, so it seemed to
me that I was bound to administer an earnest expostulation, so that
she might recover from her erring fancy. So thinking, I took my seat
in the tent. As soon as I did so, she brought me some mushrooms
she had collected for me in the morning, for she said: “You seemed
to be very fond of them.” I thanked her for her kindness, took all the
mushrooms and a cup of baked flour, and then set myself to read
my books. The girl stopped me, saying that she had something
which she must tell me, for she had heard something which filled
her with fear. Then she narrated what one of her uncles, that is one
of my male companions, had said about his intention to force me to
marry his niece. When she had concluded her story, I told her with
the greatest composure that I should be rather glad than afraid to
be killed by the brothers of her father. I had finished my pilgrimage,
I added; I had nothing to desire in this world, and I was not in the
least afraid to die. Moreover, I continued, I would not harbor any ill-
will, even if I should be killed now by her father and uncles. I should
rather thank them for hastening my departure to the plane of
Boḍhisaṭṭvas; so I would pray for them when I was enabled to reach
that Happy Abode. I would therefore ask to be killed that very
evening. The girl seemed surprised to find her revelation producing
an effect quite the reverse of what she had expected. She tried to
remonstrate with me on what she considered a foolish resolution,
and spoke some commonplaces about death and the pleasures of
life. Of course I easily refuted them, and at last she gave up the
evidently useless task of persuading me.
 QUARREL BETWEEN BROTHERS.

About four o’clock that afternoon the four returned. They must have
listened for some time to the conversation between Dawa and
myself, for as soon as they entered the tent, the most wicked of the
three brothers severely scolded Dawa for flirting with a man. Upon
this, the girl’s father at once took her side, and snappishly told his
brother that his Dawa had a father to protect her, and therefore
wanted nobody to meddle with her, much less an uncle who had
never given her even so much as one bowl of flour since she was
born.
The quarrel waxed hotter and fiercer, and the brothers began to
abuse each other and to divulge each other’s crimes. One accused
the other of being a robber, and of having murdered men at such
and such places, and was met with the recriminating accusation of
having attempted to rob the Government and of having fled for fear
of arrest. The wordy warfare at last developed into actual blows, and
the brothers exchanged fisticuffs, and even began to hurl stones at
each other. I thought I must interfere, and so I jumped up and
attempted to hold back the youngest brother as he was about to
spring at Dawa’s father. The fellow struck my cheek with his bony
knuckles with such force that I fell, and my whole frame shook with
pain. The confusion in the tent had reached its climax, and Dawa
was beginning to cry and so was also her aunt. I remained a passive
spectator of the rest of this terrible scene, for I had to lie prostrate
from the pain. Presently the sun set and the quarrel too spent itself
and the night passed without any further outbreak.
The next morning the party broke up, for each brother wanted to go
his way, the eldest with his wife, the second with his daughter, and
the third alone, as was also the case with me; so we had to
disperse, each for his own destination. One thing that troubled me
was the lack of sheep to carry my effects. At last I purchased two at
six tanka each, and separating myself from the rest proceeded in the
south-easterly direction. One of the brothers started for the north,
while I could see the others were retracing the road we had come
along.
I had heard before that I must push on rapidly, but I purposely took
the south-easterly direction, in order to throw off the scent any of
the brothers who might come after me to rob me, or even worse.
And so I proceeded in this direction, and by about sunset I reached
the brow of a hill, where I was obliged to bivouac in the open, and
on a snow-covered plain. The change was too sudden after having
lived for so long in the tent, and I could not snatch even one wink of
sleep during the night. On the following day, still continuing in the
same direction, I reached a small monastery of the name of Sha
Chen Khangba, where I remained that day and the next. For the first
time since I parted with the brothers and the troublesome women, I
felt safe, for I concluded that I was no longer in danger of being
pursued by one of the murderous gang. I saw only two priests in the
temple, and I spent most of my time in stitching my worn-out boots
and clothes.
While I was staying in the monastery one of my sheep suddenly fell
ill and died. I felt sorely grieved at his death, and read a suitable
service for him. The other sheep I had to sell, at half the price I had
paid for him, to one of four traders who arrived at the monastery
soon after I had reached it, for I could hardly manage him now, as
he was so peevish and disconsolate at the loss of his partner. To the
four men I also gave the flesh of the dead sheep, and they accepted
it with thanks. It happened that the party was travelling in the same
direction as myself, and they proposed that I should go with them.
This was quite a welcome suggestion, especially as the men were
kind enough to offer their services to carry my effects, for they had
with them a number of yaks.
So once more I had travelling companions, and I left the monastery
with a far more cheerful heart than I had when I reached it. We
proceeded in a south-easterly direction, and soon came to a small
round pond, a little over half a mile in circumference. Proceeding
along the right side of the pond, we next came to a lake which is
very long from north-west to south-east, but very narrow. The whole
circumference is said to be about forty miles. This lake is bounded
by rocky hills on all sides, and the blackish rocks scattered here and
there were partially covered, especially in the crevices and sheltered
spots between the adjoining rocks, with a thin layer of snow, so that
they presented quite a pretty sight. I ascended a small elevation
close by the lake, and looked down on it and also on the small pond.
From that height the serpentine lake looked just like the fabulous
dragon in the act of clutching a round gem, the pond corresponding
to the gem. The snow-streaked rocks were not unlike a white fleece
of cloud. This lake is known by the name of Kong-gyu-i Tso, as I
heard from my fellow travellers. After proceeding about seventeen
miles south-eastwards, with the lake on our left, we reached its
extremity. Here we were to bivouac, as we had no tent, but I could
not sleep on the snow-covered plain. I therefore passed the night in
my usual style, that is to say, in religious meditation, the best
expedient for a sleepless night.
Our road lay next day over a steep hill, and it was indeed such a
break-back ascent that it seemed to be trying even to the sturdy legs
and lungs of my Tibetan companions. As for myself, I was lucky
enough to get permission to ride on a yak’s back, and so I could
negotiate the ascent with no great difficulty. Descending the
opposite slope of the hill we soon reached a plain which, together
with all the adjoining country, was situated in the Kong-gyu district.
On this plain I noticed a white spot, not unlike a lake at a distance.
My companions informed me that the white thing was puto, and that
the white spot indicated the site of a lake which produced natural
soda.
When we reached the lake my companions eagerly collected the
deposit, put it in skin-bags and fastened it to the back of the yaks.
They told me that the soda was to be mixed with tea.
We then went on over several low undulating hills, and finally
reached the lower course of the river Chema Yungdung, where I had
narrowly escaped drowning a short time before. As the season was
now well advanced, the river was much shallower and we were able
to cross it with comparative ease. I indeed could do so with perfect
security, for I was carried on a yak’s back.
We were travelling all these days at the rate of about twenty-five
miles a day, and I should hardly have been able to make such good
progress had it not been for the fact that I could ride every now and
then on a yak. What distressed me most was bivouacking in the
open, where sleep was out of the question in the cold autumn nights
and on ground covered with snow. After proceeding some twenty-
five miles to the south-east, on the following day we reached the
Brahmapuṭra, known in this region as Martsan-gi-chu or Kobei-chu,
according to the districts which it traversed. The lordly river was
quite shallow and could be crossed without trouble, and I did so as
before on the yak’s back. We found some tents by the bank of the
river where we were allowed to pass a night—quite a cheering
change after so many nights of bivouacking.
It was a moonless night, but the sky was full of stars, which threw
their twinkling rays on the water of the river. The vast range of the
Himālayas was clearly silhouetted, so as to make its sharp outline
perceptible. The majestic scene inspired me with poetic fervor:

Like to the Milky Way in Heaven at night,

With stars begemmed in countless numbers decked,
The Brahmaputra flashes on the sight,
His banks, fit haunt for Gods, appear
In gorgeous splendors from the snowy height.

The following day I had to part from my companions, who were

going to a destination different from mine, and so I was again
thrown on my own wits and my own legs for continuing my journey.
After having travelled for so many days with the help of other
people, I now had to travel alone with nothing but my back on which
to carry my effects, and my journey on the following day was a
cheerless and fatiguing one. The load weighed heavily on my back,
and the time I occupied in taking rest was perhaps longer than that
spent in actual progress. At last I was so much exhausted that I
could hardly move my limbs.
 CHAPTER XXXII.
A Cheerless Prospect.
While I was taking rest in that helpless condition, I was fortunate
enough to see a Tibetan coming along my way leading a yak. When
he came to where I was sitting I greeted him, and asked him to
carry my luggage as far as he could without compromising his own
convenience, and promised him suitable pay for his trouble. He
willingly consented, and relieved my aching back of its load.
After travelling about three miles, I observed three men coming
towards us on horseback. They were fully armed, each with a gun, a
lance and a sword, and as they approached they looked like burly
men, wearing Tibetan hunting caps. I at once concluded that they
must be highwaymen, for evidently they were not pilgrims, the latter
generally travelling with a pack-horse or a yak to carry their
necessaries; nor could they be merchants, for they would travel in a
caravan, according to the fashion of the country. My companion
came to the same conclusion and began to show signs of fear. To
encounter highwaymen is not quite agreeable under any
circumstances, so I was not cheered at the thought of meeting those
three fellows, but I was not at all afraid, for I made up my mind to
surrender whatever things they wished to have out of my effects. I
simply wished to keep my life, and for this the highwaymen could
have no use. With that idea I boldly advanced and soon came face
to face with the three cavaliers. They asked me whence I came, and
when I replied that I was returning from a visit to Mount Kailāsa
they further asked me if I had not seen some traders on my way.
The traders were their friends, they continued, and they were
searching for them.
When I replied that I had not met with any such persons, the men
then said that I must be a Lama-priest, and as such they wished me
to perform some divination for them, to find out the whereabouts of
their friends. Now the meaning of their request was quite clear, they
wished to find the traders in order to assault and despoil them. For
my own part I was rather relieved when the three fellows disclosed
their intention, for I knew that highwaymen who were after traders
with rich goods would scorn the idea of robbing a poor Lama-priest
such as they took me to be. On the contrary they might offer some
donation to such a priest, if they asked him to undertake divination
for their sake. Highwaymen who do business on a large scale often
prove a source of substantial profit to Lama-priests, if a donation
coming from such quarters can be regarded as a legitimate profit.
Well, placed under such peculiar circumstances, I was obliged to
give them a ‘direction,’ and of course the direction I gave them was
the one which I judged least likely to be frequented by traders. The
highwaymen were highly pleased at my divination, thanked me, and
then galloped off in the direction I indicated. They did not offer me
any donation, however, for they said they had nothing to present to
me now.
My companion had remained at a distance all the while as a terrified
spectator of this strange transaction between the highwaymen and
myself. When they had galloped away he emerged from his hiding-
place, and asked me what I was talking about with those dreadful
men. I told him in detail what had passed between the highwaymen
and myself, and relieved his anxiety by assuring him that my
divination was a mere sham, and was really intended to mislead
them instead of giving them any probable direction.
After having walked along a river bank for about eight miles we
came to a tent which belonged to my companion, and there were
two or three others besides. That night I slept in the tent of my
guide, and I also stayed there during the following day in order to
give rest to my fatigued limbs. On the following morning (that is on
September 26th) I purchased a goat according to the advice of my
host, and packing my effects on the animal’s back I left the place.
I was soon after overtaken by a fearful snowstorm, which obstructed
my sight and blocked my progress. My Tibetan garment was
completely drenched, and I was wet to the skin. I could not
determine in which direction I should proceed, for the storm blinded
my eyes and I had lost my compass; but though I could not be
certain of the right direction I had to move on as best I could, for to
stand still was out of the question. My situation was growing more
and more desperate, and I was at my wit’s end, not knowing what to
do. As luck would have it, just at that moment I met with a
horseman. He at once noticed the plight I was in and kindly offered
me the hospitality of his tent. It was a little détour, he said, for one
going to Lhasa to go with him to his tent, but it would be dangerous
(though not probably fatal, as the season was not yet far advanced)
to pass that snowy night in the open; the cold was too severe to
expose oneself to it with safety at night. I gratefully accepted the
hospitable offer, transferred, as he bade me, a part of my goat’s load
on to the back of his horse, and then, leading the goat, followed the
horseman and soon reached his tent.
The following morning my host left quite early, and the people of the
tent, and of four or five others, broke up their camp and moved on
in the direction in which I also was to proceed for my journey
towards Lhasa. So I followed them, and trudged along the snow-
covered ground in a south-easterly direction for about fifteen miles. I
had not yet had occasion to talk with any of them, but I felt sure
that they would again extend to me their hospitality, and at least
allow me to share their tent at night, for they must see, I thought,
that it was impossible for me to sleep outside among the snow-
covered hills. In time the party made a halt, selected a suitable site
for pitching their camp, scraped off the snow, and set up their tents.
All that while I was watching the people at work, or gazing at the
surrounding scenery. When the tent-pitching was finished, I asked
the people of the tent in which I had slept the preceding night for
permission to enjoy a similar favor again. I was astonished to receive
from them a blunt refusal. Then I tried another tent, but with no
better success, and my earnest requests at the five or six other tents
were all in vain. I at last came to the only remaining tent, and I
thought that as this was my last chance I must somehow or other
persuade the inmates to admit me, whether they were willing to
receive me or not; so I begged them—they were an old dame and
her daughter—for permission to sleep in their tent, on the ground
that I should probably be frozen to death if I were to stay outside in
the snow on that cold night, and urged that they should take
compassion on me. I added that I might repay their kindness with a
suitable present of money. The old woman was not softened at all by
my appeal. On the contrary she was angry with me, saying that I
was insulting her by trying to force hospitality from her. Why had I
not tried other tents inhabited by men, and why should I be so
importunate with her alone? I was insulting her because she was a
woman, she added, and she insisted on my leaving her tent. When I
tried to protest against this merciless treatment she stood up in an
awful passion, and raising aloft the Tibetan tongs, with which she
was scraping together the kindled yak’s dung, she made as though
she would strike me.
 CHAPTER XXXIII.
At Death’s Door.
No one would take me into his tent, and I was thus quite at my wit’s
end. I retired to a distance of some dozen yards and, looking at the
four or five tents which appeared to be warm and cozy, remembered
Buḍḍha’s words: “For him who has no relationship to me, it is very
difficult to receive salvation from me.” These people were perfect
strangers to me, and therefore slept comfortably in their tents, while
I had to lie down on the cold ground, exposed to the severe winds.
But, I thought, the fact that I had asked them for a lodging might
have created a certain relationship, by means of which they might
yet be saved, and that it would not be quite in vain if I read the Holy
Texts for their salvation. Of course this was merely my duty as a
follower of Buḍḍha, whose love is universal. So I sat down on the
ground and recited the Buḍḍhist Text, with the kindest intentions.
After a while the girl whom I had lately asked for a lodging peeped
from her tent and stared at me, then hastily withdrew. Presently she
appeared a second time and, approaching me, said that she
supposed I was conjuring evil spirits to punish her and her mother
for their refusal to lodge me. This must not be done, said she. She
and her mother had now agreed that they should entertain me in
their tent, and she had been sent for me. There was something
comical in the fact that my kind intentions should be taken for
revengeful motives, and that those motives should be rewarded with
kindness. But I attributed all to the benevolence of Buḍḍha, and
thankfully accepted the girl’s invitation. A Buḍḍhist service was held
that evening.
The following morning I left the tent very early, and walked south-
east for two miles and a half in a hilly district. Quite unexpectedly,
two men rushed out from behind a rock and stopped me. As they
did not seem like robbers, though they were armed, I was simple
enough to think that they were natives of the place making a trip.
They approached close to me and asked me what I had. I replied, “I
had Buḍḍhism”. They did not understand what I said and exclaimed:
“What is that you have on your back?”
“That is my food.”
“What is that sticking out on your breast?”
“That is my silver.”
No sooner had the last answer been given, than the men seized my
sticks, and I understood at once that they were robbers. Promptly
making up my mind what to do, I said:
“You want something of me?”
“Of course!” one of them said, showing his teeth.
“Well, then, there is no use in hurrying. I will give you all you want.
Be calm, and say what you want.”
“Produce your silver first.”
I gave them my purse.
“You seem to have some valuables on your back. Let us see.”
I obeyed. They also demanded to see my bag, which was being
carried on the goat, and, after ransacking it, returned me the
Scriptures, the bed-clothes that were heavy, and a few other things
that were useless to them. They took, however, all my food, saying
that they needed it, although neither could I do without food.
ATTACKED BY ROBBERS.
It is a rule among the robbers of Tibet that, having taken all they
want, they should give their victim enough food for some three days,
provided that the latter read the Texts and ask for food. I thought I
would follow this custom, and I said that I possessed in my breast-
cloth a silver pagoda, containing relics of Buḍḍha, which Mr.
Ḍhammapāla of India had asked me to present to the Dalai Lama,
and which I did not wish to lose. The highwaymen at once wanted
to know if I could not give it to them, and I replied that if they
wanted it I would give it, but that as a layman could not keep it
properly, they must expect some misfortune as a punishment for
their sacrilege.
So saying, I produced the pagoda and invited them to open it. This
was probably more than they expected. They would not even touch
it, but asked me to place it upon their heads with my benediction. I
held the pagoda over their heads and, reciting the three Refuges
and Five Commandments of Buḍḍha, prayed that their sins might be
extinguished by the merit of Ḍharma.
Then I stood up, and was going to ask of them a few days’ rations,
when two men on horseback put in an appearance far ahead, and
before I could look round, the robbers had gathered together all that
they had seized, and made off in the opposite direction. They ran
over the mountains like hares, and it was quite out of the question
for me to give them chase. I thought, therefore, to ask the
horsemen for provisions. But for some reason or other they climbed
a mountain ahead of me, and did not come as far as where I was. I
called out to them and made signs by turning my right hand
inwards, according to the Tibetan custom. Perhaps my voice did not
reach them, or they had some business demanding urgent
attendance, for they paid no attention to me. Still I had left eight
Indian gold coins which I had kept close to my skin. My baggage
having been greatly diminished, I placed it all on my goat, and went
on with my journey. It was a steep mountain pass, and before I had
travelled eight miles it became dark. The night was spent as usual in
bivouacking in a crevice between the rocks.
The following morning I wished to take a north-easterly direction, so
as to reach a certain post-town; but having no compass, I could not
ascertain my bearings, and seem to have strayed off to the south-
east and eventually due south, instead of north-east, as I should
have done. The snow began to fall at three o’clock in the afternoon,
and I walked and walked until the evening, but met not a single
human being. I was exceedingly hungry, and so thirsty that I ate the
snow. One meal a day would have been sufficient for me, but the
absolute fasting gave me no small pain.
Darkness and hunger compelled me to stop, and I selected a hollow
in the ground as my bed, clearing it of snow. As there is always a
danger of being frozen dead when one is beset by a snow-storm in a
vast plain, I took the precaution to hold my breath, so as to
minimise the communication with the outside air, according to the
methods learned during my Buḍḍhist training. This, I think, is the
best method for bivouacking in the snow, and I soon fell asleep in
the hollow.
On waking early next morning I saw the snow had fallen to a great
depth, but the weather was fine; and when I looked around, I
thought the mountains ahead resembled the hilly district called
Nahru-ye, where I had once been in the company of some
herdsmen. Proceeding further, I found the familiar Kyang-chu river,
which I was delighted to see. Sustained by the hope of finding some
old acquaintances at Nahru-ye, I walked some five miles in that
direction. But nowhere was there any human being to be seen; there
was nothing but the snow. I was almost despairing, owing to my
extreme hunger and thirst, for I was entirely exhausted, though I
had no heavy baggage to carry. But I had to walk on and on, eating
a little snow from time to time to allay my hunger.
I thought that by travelling farther across the Kyang-chu river, I
should reach the place where Alchu Lama lived. He never wandered
far away from that place, and I might find him there; so I decided
upon travelling in that direction. I crossed the river about nine miles
above the place where I had crossed it on the previous occasion.
The water had decreased to about one-fifth of its usual amount, and
it was just freezing. I broke the ice with my sticks and crossed the
river. If the ice had been thick, the crossing of the river would have
been a very easy task, but the thin ice entailed the danger to the
traveller of being thrown into the deep current, and injured by the
ice-blocks. After many difficulties I reached the opposite bank, and
walked due south.
Then the baggage which was being carried by the goat got lost. It
contained what the robbers had left—a carpet made of sheep-skin,
shoes, drugs, and such things. I searched everywhere, but in vain. I
had to give up my search and proceed further, for I wished to reach
a tent before night, as sleeping amid the snow on the open field for
several nights consecutively would mean the end of my life; so I
pushed on until eight o’clock and had covered twenty miles, when
another trouble cropped up in the shape of terrible pain in the eyes,
the result of the strong glare of the sun on the snow. My eyes felt as
if they would burst, and I could not remain quiet. Moreover the snow
recommenced falling in the evening, and the cold was extreme, and
when I lay down I felt the biting coldness of the snow on my head. I
pressed the snow on my eyes, but it did not lessen the pain in the
least. A cold sweat broke out all over me from the pain and cold,
and, in trying to calm myself, I found that my body was becoming
benumbed by the frost. I tried keeping my eyes shut, and anointed
them abundantly with the oil of cloves. But slumber was far from
me. I rivetted my thoughts on Buḍḍhism, and was doing my best to
keep down the pain, when, quite unexpectedly, I was inspired with
an uta, which runs:

Upon these plains of snow, my bed is snow,

My pillow, snow; my food also the same;
And this my snowy journey, full of pain.

The effusion soothed my heart, and I felt more than ever thankful
for the beauty of the Japanese language.
The next day, October 1, 1900, at about six in the morning I decided
to proceed on my journey. The snow had ceased, and the sun was
shining brightly, to the increased pain of my eyes. I could not walk
with my eyes shut; and yet the pain of keeping them open, however
slightly, was more than I could bear. I was so overcome by it that I
would from time to time fall down, wherever I might be. I had had
no food for nearly four days, and was so weak that the smallest
stone lying in the snow would bring me down. Fortunately I
sustained no injury, owing to the softness of the snow and the
lightness of my body. There was a time, however, when I got quite
exasperated by hunger, the pain in my eyes, and the weakness of
my legs, and sat down in the snow, feeling that I was fated to die.
Intellectually, however, death was far from my thoughts. Were there
only some means of getting rid of my bodily pains, I thought I could
walk on and on, and at last reach safety.
At this juncture a horseman put in an appearance far ahead. I
strained my eyes, though with terrible pain, and thus made out that
it was a horseman. I stood up at once and signalled him to
approach. I wished to shout but could not; the effort seemed to
choke me, and it was only after enormous exertion that I squeezed
out two feeble shouts and wildly gesticulated. The horseman, having
apparently observed me, galloped towards me, to my great joy.
Soon he was beside me, asking me what I was doing in such a
desert of snow, and I told him with uncommon difficulty that I had
been robbed of most of my baggage, had lost what remained to me
en route, and had had nothing to eat for over three or four days. He
was a young man, full of sympathy. Though he was provided with
extra provisions, he said, he would give me only some sweetmeats,
made of cream and brown sugar, a food which is esteemed as a rare
delicacy in the northern steppes of Tibet. I swallowed down the food
which he gave me so hurriedly that I did not even taste it.
I then enquired of him if I could not find a lodging hereabouts. His
reply was that he was a pilgrim, and that his parents and others
were staying beside the mountain ahead of us, and that I should be
able to obtain some accommodation there. He therefore advised me
to come to his tent, and, saying that he was in hurry, galloped away
in that direction.
The distance was only a little above two miles, but I do not
remember how often I stumbled and fell down, and rested, and ate
snow, before I reached the tent. More than three hours were
occupied on the journey, and I did not reach the tent till past eleven
o’clock, when the young man came out to welcome me. His parents
congratulated me on my narrow escape from death, and entertained
me with the best sort of Tibetan food, which consisted of boiled rice
covered with butter, and accompanied by sugar and raisins. I did not
take much of the food, for fear that the sudden repletion might
injure me, but I took a little milk after a very modest repast. The
pain in my eyes was no better. There was no medicine, and the best
I could do was to cool them with snow. In spite of the fine bed with
which I was accommodated I could not sleep that night, owing to
the pain I felt in my eyes.
These people, being pilgrims, were intending to move on day by day.
The next morning, therefore, I also had to proceed on my journey.
But it was some time before they could start, for they had to pull
down the tents and pack them on the yaks. I finished my tea
therefore, and went out of doors, while they were busily engaged in
packing their effects. I had walked to the further end of a row of
four or five tents, when seven or eight ferocious Tibetan dogs
attacked me, barking loudly. Handicapped as I was with the pain in
my eyes, I could not deal with these dogs so deftly as at other
times. At first, I kept my eyes open and brandished my two sticks,
driving back the animals, which attacked me from all sides. But once
I was obliged to close my eyes, and immediately a dog behind me
seized one of my sticks. The next moment another dog fastened his
teeth on my right leg, and threw me down.
I uttered a feeble cry for help, which brought several men on the
scene, and they drove away the dogs with stones. But the blood
flowed out abundantly from the wound, which I held fast with my
hands, and I lay motionless until an aged dame brought me some
medicine, which she said was a marvellous cure for such wounds. I
dressed the wound with the medicine and bandaged it, and
attempted to rise, but in vain. It was impossible for me to stand up.
But as it was equally impossible that I should lie down there for ever,
I asked the people what they would advise me to do, and if they did
not know the whereabouts of Alchu Lama, whom I thought to be in
that vicinity. They asked me if I was acquainted with Alchu Lama,
and, on being answered in the affirmative, one of them volunteered
to carry me on his horse to the tent of Alchu Lama, who he said,
being a physician, would be able to cure me alike of the wound and
of the eye-disease. I rose with the support of the sticks, one of
which broke under me and had to be thrown away, and mounted on
the horse.
Arriving at a place where there stood two tents, I perceived that
these tents were smaller than those of Alchu Lama. Though
wondering at heart, I alighted from the horse, and enquired at one
of the tents for the Lama, and I was informed that this was not the
Lama’s tent, but that of his wife’s father. I wanted to reach the
Lama’s tents somehow, and was speaking to that effect, when the
wife, hearing my voice, said that I was the revered Lama who had
made a pilgrimage to the snowy peak of Tise, and came out to see
me.
“Where is your Lama?” I asked.
“He lives about two miles east of this place.”
“I wish to find him. Have you no one to take me to him?”
“I have nothing to do with the Lama any more, nor can I take you.
But if you want to go there, I will direct the man who has brought
you here to accompany you.”
“But why do you not yourself return to your own home?”
“Oh, there is no man so wicked as he; I intend to leave him.”
“That is not good,” said I.
Then we had a long talk, and after I had been given a repast, I rode
to the Lama’s tent.
The Lama being out, I was received by his domestics. When he
returned home in the evening, I related my adventures to him and
asked him for some medicine. He kindly dressed my wound with
excellent drugs, and gave me purgatives, saying that it was
necessary for me to purge my body in order to prevent the diffusion
throughout my system of poison which some of the dogs injected by
their bites. He also said that I should stay with him for at least a
week, in order to recuperate. Thanks to his directions, which I
obeyed, I was in a few days greatly relieved of the suffering both in
my leg and eyes.
I had experienced enough of hardship, and had very poor prospects
of an easy life in the future. But still there was a genuine pleasure in
pushing on through hardships. About that time I composed a poem:

All bitter hardships in this world of woe,

Have I thus tasted now during this life;
None will be left for me to suffer more.

One day I asked the Lama why he had sent his wife to her parents,
whereupon he explained the shortcomings of his wife. Both had their
reasons, and I could not say which was wrong. But, I said, the man
ought to have magnanimity and to console his wife, so that it was
advisable for the Lama to send for his wife to come back. I
supported my advice by the doctrines of Buḍḍhism, and made him
yield to my proposal. He sent two of his men for his wife, who, after
making some fuss, returned to his tent the same evening.
The following day, when I referred to the Discourse on the Five
Vices, which is included in the Gospel of the Buḍḍha Life Eternal
(one of the three books of the Jōdō Sect, but not found in the
Tibetan Canon) the Lama expressed his desire to hear a lecture on
it. I consented to the request and expounded the discourse on the
days following. The sermon in question treats of all imaginable vices
and sins devised by mankind, arranging them into five classes and
explaining them in the most appropriate manner. During my lecture
on this discourse the couple were so deeply moved to repentance for
their sins that they wept and at times asked me to suspend the
lecture. As their repentance was sincere, I congratulated them on
their progress in virtue. I stayed with them for some ten days, and
my bodily troubles were so much relieved that I was able to regale
my eyes with the magnificent view of the snow and ice, lit up by the
serene moon-light. This lovely scenery of nature caused me to think
of my country, and I had occasion to compose many utas, two of
which were as follows:

The spotless sky is bathed in light serene

By that cold moon with her all-tranquil ray;
This pleasant scene fires me with memories sweet
Of that dear mother-land now far away.

Here on these lonely steppes the grass is dry,

No reeds, no autumn flowers show their smiles;
On high the moon shines on these wilds alone,
Enhancing thus the loneliness profound.
THE COLD MOON REFLECTED ON
THE ICE.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookmasss.com