Beginner's Tutorial to GRASS GIS in Python

Introduction & Purpose

Throughout my brief years of study at Carleton University in the Geomatics discipline, I have noticed a shift towards a more ‘user-friendly’ focus in developing GIS tools. A great example of this initiative is the recent updates done to GRASS GIS with GRASS GIS83 improving the user GUI exceptionally with a single panel housing all the tools rather than many separate windows.

This apparent shift away from the traditional coding roots of GIS to a GUI dominant interface comes with upsides, downsides as well as an entirely new range of audience members. The focus of this tutorial is to explore the simple use of Python scripts written in Python 83 (8.3.10) using an open source IDE, in GRASS GIS83 using the pre-installed grass.script package in GRASS. The tutorial will consist of a few simple analyses using data from a few sources such as Open Ottawa and Ontario Geohub to demonstrate the capabilities of using GRASS with script rather than individual tools and the time it saves a user, especially when completing repetitive tasks.

From my personal experience the hardest part to kick-start your GIS coding experience is finding the resources and knowing where to start. This tutorial aims to fill the gaps in knowledge of a beginner GIS user looking to use code in GIS through a user friendly method available in GRASS. Therefore, the audience to this tutorial is marketed at beginner users that know the basic working principles of GIS, and are looking for a taste of scripting in there analysis process.

Software, Data and the Environment

There are a few software downloads and installations required to follow along, as well as a few preliminary steps to setting up the environment, all of which are open source and free to use! Note that this tutorial is performed on a Windows operating system and thus will follow with Windows examples, for the most part the code in BASH for Mac or Linux is similar but feel free to reference online material to compensate!

1. Download a Spyder compatible version of Python

The first step is to download a Python Interpreter off of the Python website ([1]), as well as set up the directories we wish to work with. Scroll to the bottom of the Python website and, for windows users, download the 64-bit windows installer. For Mac users install the corresponding Mac installer listed in the table.

Pay close attention to the option to add Python to your system's or user PATH. This allows your computer to search python for any executable files required. In addition, note down the path to where python was installed to be sure we can select the correct interpreter in the IDE.

2. Creating the Virtual Environment

To complete the seemingly daunting task of creating a virtual environment we are going to make use of the python package Virtualenv. This package allows the user to create a user defined environment to keep all their code related projects organized and in check.

Command Line Example

• First: Take the noted location of your Python installation, and open up either Bash Terminal or Windows Command Prompt and type the following code into the terminal switching the directories to the ones that correspond with your machine: cd C:\Users\colli\AppData\Local\Programs\Python\Python38\Scripts.

This line of code tells your system exactly which directory you want it to work from for the first step which is installing Virtualenv. Inside the Scripts folder is the application called pip which is Python's package installer. Once you have entered the line of code starting with cd, enter: pip install virtualenv, which will initiate the download and installation.

An example of the code for creating the virtual environment

• Next: We are going to create our Virtual Environment with a meaningful name related to the project as well as a target directory to create our environment. To do so, enter this line of code in respect to your specific path: virtualenv C:\Users\colli\PyGIS. Note I named my virtual environment PyGIS so I know in the future that is my environment for working with GIS related data in Python!

• Moving Forward: The next step is to enter the file explorer, and within your virtual environment folder, under the Lib folder create a folder named packages. We now need to download our Integrated Development Environment called Spyder from the pip installer.

I highly recommend the Spyder IDE ([2]) as it comes equipped with a few useful tools to beginners such as a syntax dictionary/troubleshooter as well as different panels for viewing plots and other outputs. This can be particularly useful when doing higher level analysis to data that is associated with a location, such as Census data. With the plot tab you can view any graphs created or plotted, as well as any mapping products created or visualized within the IDE, but for the purpose of this tutorial we are strictly making use of GRASS GIS to exemplify the 'foundation' GIS finds in coding.

The code for the command line is: pip install spyder --target C:\Users\colli\PyGIS\Packages with of course your path reflecting your directories.

• In Continuation: We must now activate our virtual environments and subsequently launch Spyder! There are a few options to going about this task, but for me personally as a beginner using the direct cd command to launch things really gave me an idea of how a computer runs and searches for things so we will make use of that and only touch on the PATH variable.

Enter this code to manually search the virtual environment code and activate the virtual environment: cd C:\Users\colli\PyGIS\Scripts & activate.

An example of activating an environment

• Finally: We will launch spyder which is stored in our packages folder under the bin sub-folder. We will access it using the cd command as always cd C:\Users\colli\PyGIS\Packages\Lib\bin & spyder.

This command should launch spyder from your virtual environment and is a relatively routine baseline standard to coding in an organized manner. Not all compupters will respond well to this, especially those in already poor conditions such as mine. If that is the case for you, you can always skip the virtual environment and either download spyder directly from their website ([3]).