McmcChains-class: An object to hold estimated paraeters.
In CNPBayes: Bayesian mixture models for copy number polymorphisms

An object of this class holds estimates of each parameter at each iteration of the MCMC simulation.

theta: means of each batch and component
sigma2: variances of each batch and component
pi: mixture probabilities
mu: overall mean in a marginal. In batch model, averaged across batches
tau2: overall variance in a marginal model. In a batch model, weighted average by precision across batches.
nu.0: shape parameter for sigma.2 distribution
sigma2.0: rate parameter for sigma.2 distribution
logprior: log likelihood of prior.
loglik: log likelihood.
zfreq: table of z.
predictive: posterior predictive distribution
zstar: needed for plotting posterior predictive distribution
k: integer specifying number of components
iter: integer specifying number of MCMC simulations
B: integer specifying number of batches