Add script to calculate reweighted evidence

I gave up on using the hyper-parameter posteriors for the full population and switched to using just those obtained without GW230529 to make it easier to calculate the one-left-out PPD. I am not properly normalizing the PE prior, mostly because I am lazy and don't want to do extra integrals to incorporate the cuts in chirp mass and mass ratio. I am happy to implement that if someone else calculates it. Otherwise, we can just compare the population models against each other.

@amanda.farah @anarya.ray @michael.zevin

requested review from @shanika.galaudage

requested review from @ethan.payne

added 1 commit

• 2afa140a - Add macro writing

Compare with previous version

After reviewing the code, I'm happy to sign-off on this implementation of the evidence reweighting calculation for the nsbh_pop population model (git hash 2afa140a). This calculation follows Eq 5 from https://arxiv.org/pdf/1905.05477, which is quite straightforward.

marked this merge request as ready

added 1 commit

• 95fbf0d7 - Change to log10

Compare with previous version

This last change following Ethan's sign-off just converts the final value being printed from natural log to log-10, so does not affect the code that was signed-off.

mentioned in commit 9e003db5

merged