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

Simulation of isotropic Potts model with local interaction by Gibbs sampling. Slightly faster compared to simulPotts_cpp.

1	simulPottsFast_cpp(W_i, W_p, W_x, site_order, sample, rho, n, p, iter_nb)

`W_i`	the 0-based row numbers for each non-zero element in the sparse neighbourhood matrix. integer vector.
`W_p`	the pointers to the initial (zero-based) index of elements in the colum of the sparse neighbourhood matrix. integer vector.
`W_x`	the non-zero elements of the matrix. numeric vector.
`site_order`	a specific order to go all over the sites. An integer vector or a negative number to indicate no specific order.
`sample`	A matrix containing the initial observations (by rows) for the various groups (by columns).
`rho`	A numeric corresponding to the Potts model parameter (i.e. intensity of the neighbourhood correlations).
`n`	The number of observations. integer.
`p`	The number of groups or colors. integer.
`iter_nb`	The maximum number of iteration of the Gibbs sampler. integer.