Installing Python + GIS

How to start doing GIS with Python on your own computer?

First step is to install Python and necessary Python modules that are needed to perform various GIS-tasks. The purpose of this page is to help you out installing Python and various useful GIS modules into your own computer. Even though it is possible to install Python from their homepage, we highly recommend using Anaconda which is an open source distribution of the Python and R programming languages for large-scale data processing, predictive analytics, and scientific computing, that aims to simplify package management and deployment. In short, it makes life much easier when installing new tools on your Python to play with.

Install Python + GIS on Windows

Following steps have been tested to work on Windows 7 and 10 with Anaconda3 64 bit, using conda v4.3.29 (30th October 2017).

Download Anaconda installer (64 bit) for Windows.

Install Anaconda to your computer by double clicking the installer and install it into a directory you want (needs admin rights). Install it to all users and use default settings.

Note

Note for University of Helsinki workers: you need to set the installation location as C:\HYapp so that it can be used easily by anyone without the need to pass admin credentials all the time. If you don’t have C:\HYapp -folder, create one with admin rights.

Test that the Anaconda´s package manage called conda works by opening a command prompt as a admin user and running command conda --version.

Install GIS related packages with conda (and pip) by running in command prompt following commands. Many of the GIS packages are bundled with geopandas, but there are a few that requires separate installation. As you can see below, many of the GIS packages are available from specific channel from conda called conda-forge.

# Install Geopandas
conda install -c conda-forge geopandas

# Install geoplot
conda install -c conda-forge geoplot

# Install osmnx
conda install -c conda-forge osmnx

# Install pysal
conda install -c conda-forge pysal

# Install contextily
conda install -c conda-forge contextily

# Install rasterio
conda install -c conda-forge rasterio

# Install rasterstats
conda install -c conda-forge rasterstats

# Install pycrs
pip install git+https://github.com/mullenkamp/PyCRS

# Install Dash using Pip
pip install dash==0.19.0  # The core dash backend
pip install dash-renderer==0.11.1  # The dash front-end
pip install dash-html-components==0.8.0  # HTML components
pip install dash-core-components==0.14.0  # Supercharged components
pip install plotly --upgrade  # Plotly graphing library

Test that everything works

You can test that the installations have worked by running following commands in your IPython console (comes with mini-conda).

import geopandas as gpd
import pysal
import cartopy
import geoplot
import osmnx
import folium
import dash
import rasterio
import osmnx
import contextily

If you don’t receive any errors, everything should be working!

Hint

It is also possible to install all these packages at once by taking advantage of .yml environment file that is provided by us. Using them requires a few special tricks, read more from here.

Install Python + GIS on Linux / Mac

Install Anaconda 3 and add it to system path

# Download and install Anaconda
sudo wget https://repo.continuum.io/archive/Anaconda3-5.3.0-Linux-x86_64.sh
sudo bash Anaconda3-5.3.0-Linux-x86_64.sh

# Add Anaconda installation permanently to PATH variable
nano ~/.bashrc

# Add following line at the end of the file and save (EDIT ACCORDING YOUR INSTALLATION PATH)
export PATH=$PATH:/PATH_TO_ANACONDA/anaconda3/bin:/PATH_TO_ANACONDA/anaconda3/lib/python3.7/site-packages

Install Python packages

Install GIS related packages with conda (and pip) by running in command prompt following commands (in the same order as they are listed):

# Install Geopandas
conda install -c conda-forge geopandas

# Install geoplot
conda install -c conda-forge geoplot

# Install osmnx
conda install -c conda-forge osmnx

# Install pysal
conda install -c conda-forge pysal

# Install contextily
conda install -c conda-forge contextily

# Install rasterio
conda install -c conda-forge rasterio

# Install rasterstats
conda install -c conda-forge rasterstats

# Install pycrs
pip install pycrs

# Install Dash using Pip
pip install dash==0.19.0  # The core dash backend
pip install dash-renderer==0.11.1  # The dash front-end
pip install dash-html-components==0.8.0  # HTML components
pip install dash-core-components==0.14.0  # Supercharged components
pip install plotly --upgrade  # Plotly graphing library

How to find out which conda -command to use when installing a package?

The easiest way

The first thing to try when installing a new module X is to run in a command prompt (as admin) following command (here we try to install a hypothetical module called X)

In most cases this approach works but sometimes you get errors like (example when installing a module called shapely):

Okey, so conda couldn’t find the shapely module from the typical channel it uses for downloading the module.

Alternative way to install if typical doesn’t work

How to find a way to install a module if it cannot be installed on a typical way? Well, the answer is the same is in many other cases nowadays, Google it!

Let’s find our way to install the Shapely module by typing following query to Google:

course-info/img/google_query_conda.PNG

Okey, we have different pages showing how to install Shapely using conda package manager.

Which one of them is the correct one to use?

We need to check the operating system banners and if you find a logo of the operating system of your computer, that is the one to use! Thus, in our case the first page that Google gives does not work in Windows but the second one does, as it has Windows logo on it:

course-info/img/conda_shapely_windows.PNG

From here we can get the correct installation command for conda and it works!

course-info/img/install_shapely.PNG

You can follow these steps similarly for all of the other Python modules that you are interested to install.