MCMC

Ongoing development towards MCMC improvements

Presentation on MCMC options

Cole Monnahan, Jim Thorson, and Ian Taylor gave a talk (Feb-27-2013) detailing options for MCMC in ADMB, and when and how they should be used. Shows the mcrb and mcprobe algorithms, as well as demonstrating the MCMCMC and hybrid methods.

Updating the MCMC documentation

Ideas for improvements/clarifications for the existing documentation on MCMC methods.

MCMC Guide for ADMB

The goal of this guide is to outline and describe the steps needed to conduct a Bayesian analysis in AD Model Builder. Included are general descriptions of Bayesian inference, priors, work flow and two built-in MCMC algorithms. We hope that the guide will enable users to take full advantage of the features of ADMB that are built in, but not yet well documented.