search

grambank / grambank-analysed

3 stars 0 forks source link

added normalising constant to marginal likelihoods and calculated Bay… #46

Closed rdinnager closed 2 years ago

rdinnager commented 2 years ago

…es Factors between best model and each other model per feature. Did for both gaussian and integration. I think integration should be preferred. Also fixed cpo summary. I don't think we can use the PIT scores.

HedvigS commented 2 years ago

This makes sense to me. you approve @SamPassmore ?

HedvigS commented 2 years ago

(I'm very glad that we're doing this in a separate script by the way, saves me re-running things :) )

HedvigS commented 2 years ago

@SamPassmore if you're happy with these changes I'll merge this on

HedvigS commented 2 years ago

i tried to resolve the conflicts with this branch and main, i hope i got it right.

rdinnager commented 2 years ago

Looks fine to me.