[adegenet-forum] bug in fstat and genind2hierfstat

[Jombart, Thibaut]

Dear list, 

a bug reported by Patrik Mraz a few days ago pointed out some conversion issues that occurred when using genind2hierfstat, and thus also when using the fstat function. 
The error came from the unexpected way data are interpreted by hierfstat. Bias in the results of fstat would have occurred whenever some loci had more than 10 alleles.

The problem is all fixed now. I posted a patch on the adegenet website (section download), which should be available within a couple of hours.

Sorry for the inconvenient.

Best regards

Thibaut 

note: the function pairwise.fst, which does not use hierfstat, is not affected