Comments on "A theoretical analysis of Monte Carlo algorithms for the simulation of Gibbs random field images" [with reply]

The commenter contends that, contrary to the conclusions of the above-titled paper by J.K. Goutsias (see ibid., vol.37, no.6, p.1618-28, Nov. 1991), the relative entropy of the initial distribution for a Monte Carlo simulation is not monotonically related to the degree of convergence after some given number of steps, and optimal single-step convergence is not achieved by the Gibbs sampler with systematic site ordering. In his reply, Goutsias explains the intent of his work, but expresses agreement with many of the comments. He provides a corrected statement of Theorem 3.<
>