Applied Bayesian Inference in R using MCMCpack

Abstract

Over the past 15 years or so, data analysts have be- come increasingly aware of the possibilities afforded by Markov chain Monte Carlo (MCMC) methods. This is particularly the case for researchers interested in performing Bayesian inference. Here, MCMC methods provide a fairly straightforward way for one to take a random sample approximately from a posterior distribution. Such samples can be used to summarize any aspect of the posterior distribution of a statistical model. While MCMC methods are extremely powerful and have a wide range of applicability, they are not as widely used as one might guess. At least part of the reason for this is the gap be- tween the type of software that many applied users would like to have for fitting models via MCMC and the software that is currently available. MCMCpack (Martin and Quinn, 2005) is an R package designed to help bridge this gap.