rajanil / fastStructure

A variational framework for inferring population structure from SNP genotype data.
MIT License
134 stars 50 forks source link

Marginal likelihood = nan #40

Open nadjenajar opened 6 years ago

nadjenajar commented 6 years ago

After running my data set (358 SNPs) with a simple prior to k=15 and a logistic prior to k=4, I'm now getting this message as my marginal likelihood for k=5 and k=6:

... 360 -0.3007400419 0.0000752459 326.359 600 -0.2987151228 0.0000319624 323.266 170 nan nan 306.530 Marginal Likelihood = nan Total time = 4977.0717 seconds Total iterations = 170

I don't know what this means - any suggestions as to how to fix or what's causing this?

Thanks so much,

Nadje