What are the best ways, if any, to test for the convergence (or pseudo-convergence) of the Markov chain to the equilibrium distribution?