Persistable: persistent and stable clustering

Persistable is an implementation of density-based clustering algorithms intended for exploratory data analysis.What distinguishes Persistable from other clustering software is its visualization capabilities.Persistable's interactive mode lets the user visualize multi-scale and multi-density cluster structure present in the data.This is used to guide the choice of parameters that lead to the final clustering.Persistable is based on multi-parameter persistence (Botnan & Lesnick, 2022), a method from topological data analysis; the theory behind Persistable is developed in (Rolle & Scoccola, 2020).Persistable is implemented in Python, with the most expensive computations-in particular the implementations borrowed from (Virtanen et al., 2020) and from the high-performance algorithms for density-based clustering developed in (McInnes & Healy, 2017) and implemented in (McInnes et al., 2017)-done in Cython.Persistable's interactive mode is inspired by RIVET (The RIVET Developers, 2020) and is implemented in Plotly Dash (Plotly Technologies Inc., 2015).We test the core algorithms as well as the graphical user interface in Ubuntu, macOS, and Windows.We have designed Persistable with the goal of making both its computational components and its GUI easy to extend.We hope to keep adding high performance topological inference functionality.The documentation for Persistable can be found at persistable.readthedocs.io.