PIVMET: pivotal methods for Bayesian relabelling in finite mixture models

The identification of groups’ prototypes, i.e. elements of a dataset that are representative of the group they belong to, is relevant to the tasks of clustering, classification and mixture modeling. The R package pivmet includes different methods for extracting pivotal units from a dataset, to be exploited for a Markov Chain Monte Carlo (MCMC) relabelling technique for dealing with label switching in Bayesian estimation of mixture models. Moreover, consensus clustering based on pivotal units may improve classical algorithms (e.g. k-means) by means of a careful seeding.