sfMM-simulPotts_cpp: Potts model simulation
In bozenne/MRIaggr: Management, Display, and Processing of Medical Imaging Data

C++ function that simulates an isotropic Potts model by Gibbs sampling. Experimental.

1 2	simulPotts_cpp(W_SR, W_LR, sample, coords, site_order, rho, distance_ref, iter_max, cv_criterion, regional, verbose)

`W_SR`	the local neighbourhood matrix. dgCMatrix. Should be normalized by row (i.e. `rowSums(W_SR)=1`).
`W_LR`	the regional neighbourhood matrix. dgCMatrix. Should contain the categroy of distance.
`sample`	the initial observations (by rows) for the various groups (by columns). matrix.
`coords`	the voxels coordinates. matrix.
`site_order`	a specific order to go all over the sites. An integer vector or a negative number to indicate no specific order.
`rho`	a numeric corresponding to the Potts model parameter (i.e. intensity of the neighbourhood correlations).
`distance_ref`	the distance defining the several neighbourhood orders relatively to `Wdist_LR`. numeric vector.
`iter_max`	the maximum number of iteration of the Gibbs sampler. integer.
`cv_criterion`	if the maximum absolute difference of the probabilistic membership between two iterations is below `critere` then the Gibbs sampler is stopped. numeric.
`regional`	should the regional potential be computed ?. logical.
`verbose`	should the convergence criterion be printed at each step ? logical.