Posterior analysis (convergence diagnostics)

[dal]

After simulating two parallel MCMC chains, dynare provides the Brooks and Gelman (1998) convergence monitoring diagnostics, which provide a necessary (not sufficient) condition for convergence. In addition to that, I would be interested in comparing the estimated distribution of the two chains. The two chains may not have converged to the same distribution, even if BG diagnostics do not deny convergence. Thus, this type of analysis would help to have a stronger confidence about the success of the estimation. Is there an easy way to plot the distribution of each of the parallel chains, for each parameter, in the same graph?