Bayesian analysis of CGH microarrays fitting Hidden Markov Chain models. The selection of the number of states is made via their posterior probability computed by Reversible Jump Markov Chain Monte Carlo Methods. Also returns probabilistic common regions for gains/losses.