A Review of the BMS Package for R

Shahram Amini, Virginia Polytechnic Institute and State University
christopher parmeter, University of Miami

Abstract

This paper describes the relative merits and attractiveness of the newest Bayesian model averaging package, BMS, available in the statistical software R to implement a Bayesian model averaging exercise. This package provides the user a wide range of customizable priors for conducting a BMA analysis, provides ample graphs to visualize results and offers several alternative model search mechanisms.