Bayesian data analysis using R

Bayesian data analysis includes but is not limited to Bayesian inference (Gelman et al., 2003; Kerman, 2006a). Here, we take Bayesian inference to refer to posterior inference (typically, the simulation of random draws from the posterior distribution) given a fixed model and data. Bayesian data analysis takes Bayesian inference as a starting point but also includes fitting a model to different datasets, altering a model, performing inferential and predictive summaries (including prior or posterior predictive checks), and validation of the software used to fit the model. The most general programs currently available for Bayesian inference are WinBUGS (BUGS Project, 2004) and OpenBugs, which can be accessed from R using the packages R2WinBUGS (Sturtz et al., 2005) and BRugs. In addition, various R packages exist that directly fit particular Bayesian models (e.g. MCMCPack, Martin and Quinn (2005)). In this note, we describe our own entry in the “inference engine” sweepstakes but, perhaps more importantly, describe the ongoing development of some R packages that perform other aspects of Bayesian data analysis.