Running terastructure with k=1
gokceneraslan opened this issue · 2 comments
gokceneraslan commented
I'd like to use terastructure to check if there are multiple populations in a supposedly homogeneous dataset. For this reason I ran it first with k=2
and it converged nicely. But I need to compare the held-out likelihood for various values of k
, including k=1
. When I try to run it with k=1
to get a baseline likelihood value, it never stops running, although likelihood stays the same:
whao89 commented
(oops, misclicked earlier)
K=1 should correspond to just the marginal allele frequencies per SNP...let me look into the convergence.