Markov Chain Monte Carlo for Potts Models
Do Markov chain Monte Carlo (MCMC) simulation of Potts models
(Potts, 1952, ),
which are the multi-color generalization of Ising models
(so, as as special case, also simulates Ising models).
Use the Swendsen-Wang algorithm (Swendsen and Wang, 1987,
) so MCMC is fast.
Do maximum composite likelihood estimation of parameters
(Besag, 1975, ,
Lindsay, 1988, ).