Now the lower triangular expansion on the matrix of loadings is optional:

lowerTriangular: logical value indicating whether a lower triangular parameterization should be imposed on the matrix of factor loadings (if TRUE) or not. Default: TRUE.

Now the lower triangular expansion on the matrix of loadings is optional.

In this new version I have removed all dependence on the orphaned doRNG package.

Pre-print available on arXiv: Clustering Multivariate Data using Factor Analytic Bayesian Mixtures with an Unknown Number of Components

Clustering Multivariate Data using Factor Analytic Bayesian Mixtures with an Unknown Number of Components

Bug fix in plot for Lamba_map.

Bug fix in plot for Lamba_map.

Improved plot and summary methods. See CRAN page.

Improved plot and summary methods. This new version will be available to CRAN later this month.

This version fixes a bug to output which caused the main function to crash at certain cases. Not on CRAN yet.

See CRAN page

The fabMix function now features a new argument (parallelModels), allowing the user to run different models in parallel. This is combined with the pre-existing option to run heated chains in parallel, thus, parallelization is now implemented in both model-level and chain-level.

Version 3.0 introduces new parsimonious models.

Since version 4.0, the package is integrated with C++ code.

Version 4.1 improves the Lambda output: all MCMC values are now exported to a single file (instead of multiple ones, as done in previous versions).

Please use version 2.0 of the software for the relevant R scripts. But note that

1. Recent versions are recommended.
2. All package versions smaller than 4.2 are compatible only with Linux distributions.