PostProcessChain {Geneland} | R Documentation |
Computes posterior probabilities of population membership for each pixel of the spatial domain.
PostProcessChain(coordinates,genotypes,allele.numbers, path.mcmc,nxdom, nydom,burnin)
coordinates |
Spatial coordinates of individuals. A matrix with 2 columns and one line per individual. |
genotypes |
Genotypes of individuals. A matrix with one line per individual and 2 columns per locus |
allele.numbers |
A vector of integer containing the number of possible allele for each locus |
path.mcmc |
Path to output files directory |
nxdom |
Number of pixel for discretization of the spatial domain in the horizontal direction |
nydom |
Number of pixel for discretization of the spatial domain in the vertical direction |
burnin |
Number of iterations of the chain to throw away.
WARNING : this argument should be given the number of stored
iterations (and not the number of computed iterations which differ
if burnin !=1). If you have
nit =100000 and thinning =100, then only 1000 iterations
are stored. Then burnin =10 will throw away 10 stored
iterations, namely 100*10 computed iterations. |
‘path.data’ directory should contain at least three files named exactly : ‘coordinates.txt’, ‘genotypes.txt’ and ‘allele.numbers.txt’ and containing respectively the spatial coordinates, the genotypes and the number of alleles per locus.
See format of files simulated by simFmodel
for an
example.
Posterior probability of population membership for each pixel are
written in an ascii file called ‘proba.pop.membership.txt’
(one column per population,
npopmax
values are computed for each pixel,
pixels are stored row-wise starting from bottom left).
Another file called ‘proba.pop.membership.perm.txt’ tries
to get rid of label switching
issues by labelling the population according to a fixed constraint.
(This has proved to be usefull with a small number of loci, (e.g. nloc=3), for well
differentiated populations.)
Gilles Guillot
A spatial statistical model for landscape genetics, Guillot, Estoup, Mortier, Cosson, Genetics, 2005
Guillot, G., Geneland : A program for landscape genetics. Molecular Ecology Notes, submited.