This is the fourth in a sequence of posts designed to introduce econometrics students to the use of Markov Chain Monte Carlo (MCMC, or MC2) simulation methods for Bayesian inference. The first three posts can be found here, here, and here, and I'll assume that you've read them already. The emphasis throughout is on the use of the Gibbs sampler.
The first three posts took a look "inside the box", to see that the Gibbs sampler entails. I looked at some R code that could be used to show the sampler "in action". One way to think about those posts is that they were analogous to an explanation of OLS regression, with code that assembled the X matrix, the (X'X) matrix, showed you how to invert the latter matrix, and so on. It's important to understand what is going on when you select "OLS" in your favourite package, but you certainly wouldn't dream of constructing the estimator from first principles every time you wanted to use it.
It's the same in the case of our Bayesian analysis. So, the purpose of this post is to show you simple it is, in practice, to use R to estimate regression models using Bayesian methods, and to implement Bayesian Posterior Odds analysis for model selection. We can just take advantage of the R packages that have been developed already to help us.