GNU R package for Markov Chain Monte Carlo simulations

 Simulates continuous distributions of random vectors using Markov
 chain Monte Carlo (MCMC). Users specify the distribution by an R
 function that evaluates the log unnormalized density. Algorithms are
 random walk Metropolis algorithm (function metrop), simulated
 tempering (function temper), and morphometric random walk Metropolis
 (Johnson and Geyer, Annals of Statistics, 2012, function
 morph.metrop), which achieves geometric ergodicity by change of
 variable.