Mathematica Markov Chain Monte Carlo
Mathematica package containing a general-purpose Markov chain Monte Carlo routine I wrote. Includes various examples and documentation.
Features:
- Convenience wrapper for fitting models to arbitrary-dimensional data with Gaussian errors
- Handles both real-valued and discrete-valued model parameters
- Uses Metropolis algorithm with decaying exponential proposal distribution
- Progress monitor; support for auto save/resume
Files:
mcmc.m
: Package file
mcmc_demonst.nb
: Demonstrations and documentation