popsim-consortium / analysis2

Analysis for the second consortium paper.
8 stars 14 forks source link

fix scaling of mean s for grapes #102

Closed lntran26 closed 1 year ago

lntran26 commented 1 year ago

Grapes outputs the population-scaled E(s) rather than E(s). This PR fixes that by calculating Ne from Grapes' theta output and rescale E(s) with the calculated Ne for plotting.

DFE inference results ran using tiny_config.yaml before and after this grapes E(s) scaling fix:

Before: dfe inference benchmark_old

After: dfe inference benchmark_new