This repository contains a Python implementation of the Multi-restart Bicriterion Pairwise Interchange (MBPI) algorithm for the anticlustering problem. The algorithm, proposed by Brusco et al., is a bicriterion stochastic search method for finding a Pareto set of maximally dissimilar partitions across a metric space.
This implementation works for variably sized groups.
For more information on the algorithm, please refer to the original paper as well as this project's corresponding blog post for a more detailed explanation.