search

joezuntz / cosmosis

Other
22 stars 16 forks source link

Possible error in implementation of dynesty and nautilus #80

Closed johannesulf closed 1 year ago

johannesulf commented 1 year ago

I believe both dynesty and nautilus may not be correctly implemented/interfaced. The problem is that dynesty seems to be passed the posterior function when it expects the likelihood function. I then modeled the implementation of nautilus into cosmosis after dynesty's, likely copying this mistake. Sorry for not catching that earlier.

joezuntz commented 1 year ago

Many thanks for spotting this! I have released a fixed version, 2.3.2.