Python for Geographic Information System

Welcome to python for GIS training. In this course you will earn how to use python, and its libraries to implement your own GIS features.

During the corse, new materials will appear daily in this repository.

Scope

The course covers:

Shapely
Geopandas
Map projections
Geocoding
Spatial queries
- Point in Polygon
- Spatial join
- Overlay analysis
- Aggregation
Data classification
Retriving Open Street Map data
Network analysis
Visualization

We will NOT cover using ArcGIS (arcpy) + QGIS with Python. GIS universe is very diverse, and these subjects would require separate courses. For the interested ones, there are some meterials available.

You will learn how to use libraries like shapely, geopandas, mapclassify, bokeh, osmnx and networkx. By exercising, you will also strenthen your understandig of pandas and matplotlib.

To learn more about GIS stack in python, go here.

Before the course

To prepare for the course, make sure all dependencies are installed. You will need python 3.8.5, git and an IDE. You will also need to install required python packages, and jupyter notebook.

System requirements

To participate in the course, you will need:

python 3.8.5 or newer

To check your python version, run in the terminal: python --version If you need to install, or upgrade python, follow these instructions

Required packages

Now, let's install the required packages. We encourage you to use virtual environment.

In your terminal run:

python3 -m venv venv to create virtual env
. venv/bin/activate to activate it on Linux or MacOS, or venv\Scripts\activate.bat on Windows
pip install -r requirements.txt to install required packages
pip freeze to see what you are using

Geopandas on Windows If you are a Windows user, the package manager may fail installing geopandas. Geopandas require C++ dependencies, and if those are not setup correctly, the installation will fail.

If you encounter issues when installing geopandas, follow this tutorial.

Setting up jupyter notebook

Now, let's setup jupyter notebook. This will be very handy when working with map visualizations. Our recommendation is to install jupyter globally, and its kernel locally (in the venv). It allows you to have as many kernels (python versions) as projects, but only one jupyter notebook.

To setup jupyter

deactivate to quit the venv
pip install notebook to install jupyter notebook globally
. venv/bin/activate or venv\Scripts\activate.bat to activate venv again
pip install ipykernel to install kernel in venv
python -m ipykernel install --user --name=gis-training to attach kernel to your project
jupyter notebook to start a notebook

A good tutorial on configuring jupyter in virtual environment is available here

If you are using VS Code, we highly recommend this jupyter extension. It will allow you to execute notebooks directly in IDE.