[R-sig-genetics] Hierfstat: test.between() with many loci

[Kevin Weitemier]

Hello,

I'm finding some odd behavior when using hierfstat to perform  
permutations to assess the significance of population structure. I'm  
using a SNP dataset with 96 individuals and ~7300 SNPs.

When performing test.between() with a small number of SNPs, the  
process seems to work properly, but as the number of SNPs is increased  
the numbers returned in g.star start to contain repeated values,  
including zeros. With large numbers of SNPs, all of the values in  
g.star are zero.

This seems like odd behavior since I thought the values in g.star  
would increase with more SNPs, rather than go toward zero.

I'm using hierfstat v. 0.04-22, and have prepared a RMarkdown page here:

https://dl.dropboxusercontent.com/u/31360343/hierfstat_behavior.html

Thanks for any help,
Kevin Weitemier

-- 
Kevin Weitemier, M.S.
Ph.D. Candidate
Oregon State University
Dept. of Botany & Plant Pathology
2082 Cordley Hall
Corvallis, OR 97331