/mcmc

This repository is for community detection metrics such as ARI, NMI, Purity, Normalized Cut and Modularity.

Primary LanguageJupyter Notebook

mcmc

Some python code to simulate Markov chain Monte Carlo.

  1. Hard-core model (e.g., binary coloring graph)
  2. q-coloring graph with systematic sweep Gibbs sampling + Fisher Yates shuffle O(q)
  3. Basic MC simulation.

Implementation of several clustring evaluation metrics:

Normalized cut + modularity for non-groundtruth dataset. Adjusted Rand Index (ARI), Rand Index (Rand Index), Normalized Mutual Information (NMI) and Purity