This git repository contains code for $k$-means clustering for use on persistent homology representations.

The repository is structed as follows:

• PD_subsample contains code from PD-subsample, namely the ApproxPH.py script that we use and adapt code from to construct persistence measures and calculate distances between.
• src contains all code written for this project:
  • data_utils contains scripts generating the simulated data and processing the real-world data used in the project. This directory also contains methods used to clean and process persistent homology representations.
  • pd_pm_kmeans.py contains the persistence diagram and persistence measure $k$-means algorithms.
  • synthetic_data_clustering.ipynb demonstrates how the code was used to generate results on simulated data.