This repository contains a series of workbooks and datasets that are useful for learning how to calculate your own transit access to opportunity measures and distributive equity measures using only open-source data and software.
These notebooks use r5py, a Python-based wrapper for Conveyal's R5 routing engine, which is very fast and has features that make it especially good for calculating transit travel time matrices.
There are also workbooks here that focus on calculating access to opportunity metrics as well as distributive equity metrics across population groups, in order to identify potentially systemic gaps in the provision of transit service.
This repository's analysis focuses on Calgary, Canada.
It is possible to run this analysis using a cloud-hosted Jupyter notebook service called Binder (click the "open in binder" icon above to do so), however running travel time matrices requires a decent amount of computational resources and will almost certainly be more robust if run locally. Please follow the installation instructions below to get it set up on your computer.
There are may different ways to install both r5py and the other required tools on your computer. If you are familiar with environment management you can loosely follow this guide. If you are starting from scratch, here are the recommended steps:
You can get the code either by cloning this repository using Git, or by downloading a Zipped file of the code by going to Code -> Download ZIP on the main repository page. Extract this zipped folder somewhere on your computer.
If you have not installed Python or Conda on your computer before, your best bet is to start with Mambaforge (available for Windows/Linux/Mac).
Depending on how things have been installed, you may have to add Conda to your PATH variable, and then run the command:
conda init
Which will add (base) to the start of your terminal line.
If you're starting from another conda distribution, make sure you have mamba available on your system by installing it into the base environment (If you've used Mambaforge above you can skip this step):
conda install -n base -c conda-forge mamba
Next you need to install the dependencies. To do this you will need to open a command prompt or terminal and navigate to the repository folder. To ensure you have the environment you need, you should install directly from the local_install.yml file located in the main repository folder:
mamba create -n ghn -f local_install.yml
This will install a whole number of useful packages. You will need to confirm installation with y at some point. To check that your installation worked, you need to activate your environment:
conda activate ghn
And then check packages are working with
python
>>> import r5py
This might take a little while if it's the first time you're importing as r5py needs to download a file. If you see no errors, you can exit the Python interpreter with exit() and you're good to go.
You may need to install a Java Development Kit on your system (JDK), version 11. You can find downloads for this on the Oracle website. Installing this on your system should enable Java to work properly and is the first step to check if you run into issues above when importing r5py.
To work with the notebooks and to follow along in the workshop launch JupyterLab with the command:
python -m jupyter-lab
Much of the code and analysis in this repository comes directly from or is inspired by:
- The r5py usage guide
- The Equity of Transport Futures code and guidebook
If you end up using r5py in your work, please consider citing it properly