search

popsim-consortium / demes-python

Tools for describing and manipulating demographic models.
https://popsim-consortium.github.io/demes-docs/
ISC License
19 stars 6 forks source link

in_generations() is not rescaling migration rates #37

Closed apragsdale closed 4 years ago

apragsdale commented 4 years ago

Looks like the in_generations() function does not rescale migration rates (i.e. converting from years to generations does not change the migration rates from units of per-year to per-generation).

Should be a quick fix - I'll have a go at it (unless this was the intended behavior?)

grahamgower commented 4 years ago

I thought we decided that migration rates should always be per generation? #30

molpopgen commented 4 years ago

I'd like to further propose that the rate be per "gamete", per generation. In some areas, like quantitative genetics, mutation rates are often written as per offspring, and thus differ by a factor of the ploidy. I think we should have the "strict pop-gen" definition here.

apragsdale commented 4 years ago

Ok, yes now I vaguely remember we had agreed that all rates should be per-generation. And I also agree that that should be "per gamete per generation", like you suggested Kevin.

If that's the case, then I don't think we need to fix the in_generations code like the PR I opened. But some of our examples haven't been updated to be in units of per generation instead of per year, which is what threw me off. (The demes.examples.gutenkunst_ooa() model is still in per year, so I should fix those.)