bspcov: An R Package for Bayesian sparse covariance matrix estimation

The bspcov R package provides a Bayesian inference for covariance matrices. The bspcov is developed to aid in research that involves estimating constrained covariance matrices by enabling the use of state-of-the-art Bayesian inference methods. It consists of the main functions bmspcov, sbmspcov, bandPPP and thresPPP that conduct posterior inference for sparse or banded covariance matrices. The functions bmspcov and sbmspcov implement block Gibbs samplers based on beta-mixture and screened beta-mixture shrinkage priors, respectively. The functions bandPPP and thresPPP implement a direct posterior sampling from the post-processed posterior for banded and sparse covariance matrices. We demonstrate how to use the main functions with real data applications.