This repo is the companion to the FOSS4G 2025 talk of the same title. There is an explanatory web-book rendered at https://shriv.github.io/foss4g-2025/. Talk slides are in pdf form. N.B the repo is heavy due to the quarto site and a couple of input datasets. These datasets (urban/rural and suburb geometries) are available from Statistics NZ but I added them to the repo to make it easier for re-running the code. If there are any problems or you have any suggestions, please open an issue.
To render the Quarto web-book, you just need to run make setup to get the data processed and make render to generate a local version of the Quarto web-book. Note, this should work on unix operating systems like MacOS and Linux. There is a Dockerfile as well for running the code inside Docker.
The code behind the web-book largely mostly Python but also uses R for the spatial visualisations. The project uses uv to manage the Python dependencies and renv to manage R dependencies.
Note the clustering algorithms require a reasonable amount of RAM - it runs okay on an M2 Macbook Pro with 16 GB RAM. If the code fails to run (especially the Auckland file) please run it on a larger machine!