MurraySchemePotts: Simulate the parameters of a Potts model using the Murray...

In ick003/SpTMixture: Mixture models with spatio-temporal dependencies

Description

This function allows you to simulate the parameters of a Potts model using the Murray Scheme.

Usage

MurraySchemePotts(beta.init, range.beta = c(0, 1), n.colors = 2,
  Coords = NULL, Nb = NULL, col.obs, N.run = 100, print = FALSE)

Arguments

beta.init	Starting parameter values for the parameters of the Potts model.
range.beta	Range of values for the parameters of the Potts model.
n.colors	number of different "spins".
col.obs	Binary matrix of colors.
N.run	Number of MCMC runs to perform.
Nb.list	List of neighbours.

References

Murray, I. and Ghahramani, Z. and MacKay, D. (2006) MCMC for doubly-intractable distributions. UAI Proceedings 22 359–366. Everitt, R. (2012) Bayesian Parameter Estimation for Latent Markov Random Fields and Social Networks. Journal of Computational and Graphical Statistics 21 940–960.

Examples

beta = c(1.2, 1.2,1.2,1.2)
Obs.loc = expand.grid(1:5,1:5)
id.pos = sample(1:4, nrow(Obs.loc), replace = TRUE)
Res = SwendsenWangAlg(beta, Coord = as.matrix(Obs.loc), col.obs = id.pos, Nrun = 20)
par(mfrow = c(1,1))
col = grey.colors(4)
image(matrix(Res$col.new, ncol =5), col = col)
beta.init = runif(1,0,1)
set.seed(1)
beta.mcmc = MurraySchemePotts(beta.init = beta.init,Coords = Obs.loc ,n.colors = 4, col.obs = Res$col.new, N.run = 5000, range.beta = c(0,3))
Nbs = getBonds(as.matrix(Obs.loc), NN = 4, th = 2)
beta.mcmc.nb = MurraySchemePotts(beta.init = beta.init,Nb = Nbs,n.colors = 4, col.obs = Res$col.new, N.run = 5000, range.beta = c(0,3))

print(beta.mcmc$rate)
par(mfrow = c(2,2))
hist(beta.mcmc$beta.sample, breaks = 50, xlim = c(0,1.75))
plot(beta.mcmc$beta.sample, type = "l", ylim = c(0,1.75))
hist(beta.mcmc.nb$beta.sample, breaks = 50, xlim = c(0,1.75))
plot(beta.mcmc.nb$beta.sample, type = "l", ylim = c(0,1.75))

ick003/SpTMixture documentation built on May 18, 2019, 2:32 a.m.