minpts-check: one-run predictor tells if your HDBSCAN clusters survive min_samples changes
leland_mcinnes · x · 2026-11-12
A new MIT-licensed GitHub tool, minpts-check, answers whether your HDBSCAN clustering survives a small change of minsamples. It computes an shat statistic flagging at-risk clusters (those likely to fall below the next minclustersize) and provides a certificate — a sufficient condition checked from a single run — under which raising minsamples provably changes no labels outside an explicit set of points. Works with sklearn's HDBSCAN and the hdbscan package, installable via pip. Implements checks from the not-yet-peer-reviewed Coutinho (2026) preprint; retweeted by HDBSCAN co-author Leland McInnes.
More from coding & agent
- United Airlines to Ban Personal AI Agents From Booking Flights for Customers — Polymarket · 2026-11-12
- Agentic Search explained: a 4-search loop costs about 5x one search — and queries leak to more companies — adnan_hashmi · 2026-11-12
- AWS says thousands of autonomous AI agents wrote every line of AWS Emulator — SumitGup · 2026-11-12
- freeCodeCamp launches free 2-hour course on agentic coding with Codex — adnan_hashmi · 2026-11-12
- A criminally underrated resource comparing tool-calling reliability across models and providers — ricklamers · 2026-11-12
- Organizing 55k data points is easy: enterprise data prep for agents is the hard part — BenBajarin · 2026-11-12