A C++ class library for Monte Carlo Markov Chain Analysis of Bayesian models

MCMC++ is a class library intended to speed and simplify development of Markov Chain Monte Carlo models for analysis of data in a Bayesian context. It is distributed under terms of the GNU General Public License. The version currently available has been adequate for the tasks I've given it, and I believe that it's quite stable and reliable.

The documentation is, however, incomplete. Although each of the public methods is documented, I haven't had time to provide anything like an introductory tutorial. You'll have to use the simple examples provided (in examples/) as a model. You'll find HTML documentation in html/index.html and PDF documentation in latex/refman.pdf. Both are included in the distribution files and also provided here.

If you download the library and use it your work, please let me know. If you run into problems using it, I'll try to help, but don't be offended if it takes me a while to respond. This isn't the only project I have going.

And whatever you do remember (as they say in the WinBUGS documentation)

MCMC is dangerous!

And keep in mind that if MCMC in WinBUGS is dangerous,

MCMC++ is even more dangerous!

Use at your own risk.


Development of MCMC++ is supported, in part, by a grant from the National Institutes of Health (1R01-GM068449-01A1).

Kent Holsinger