emclustr is a package for finite mixture modeling via EM algorithm. Its main extensions to the mclust package are (1) mixture modeling non-Gaussian distributed data, and (2) mixture modeling for Gaussian distributed data with missing values
- Oct 2015 - v 0.4 -- wrote and passed unit tests for all functions
- Sept 2015 - v 0.3 -- updated documentation, examples, and added test script to GitHub.
- Sept 2015 - v 0.2 -- re-written to use
library(mvtnorm), add log-likelihood calculations, and calculate BIC for all models. - June 2014 - v 0.1 (alpha version)