SW: Generate Random Samples from a Compound Potts Model by the...
In PottsUtils: Utility Functions of the Potts Models

Description Usage Arguments Details Value References See Also Examples

Generate random samples from a compound Potts model using the Swendsen-Wang algorithm.

1	SW(n, nvertex, ncolor, edges, weights, beta)

`n`	number of samples.
`nvertex`	number of vertices of a graph.
`ncolor`	number of colors each vertex can take.
`edges`	edges of a graph.
`weights`	weights of edges. One for each corresponding component in `edges`. The default values are 1s for all.
`beta`	the parameter inverse temperature of the Potts model.

We use the Swendsen-Wang algorithm to generate random samples from a compound Potts model. See rPotts1 for more information on the compound Potts model.

The output is a nvertex by n matrix with the kth column being the kth sample.

Robert H. Swendsen and Jian-Sheng Wang (1987) Nonuniversal Critical Dynamics in Monte Carlo Simulations Physical Review Letters vol. 58, no. 2, 86-88

Dai Feng (2008) Bayesian Hidden Markov Normal Mixture Models with Application to MRI Tissue Classification Ph. D. Dissertation, The University of Iowa

Wolff, BlocksGibbs

  #Example 1: Generate 100 samples from a Potts model with the
  #           neighborhood structure corresponding to a
  #           second-order Markov random field defined on a 
  #           3*3 2D graph. The number of colors is 2.
  #           beta=0.1. All weights are equal to 1.
 
  edges <- getEdges(mask=matrix(1, 2, 2), neiStruc=rep(2,4))
  set.seed(100)
  SW(n=500, nvertex=4, ncolor=2, edges, beta=0.8)