You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Usually in cobaya I can include a bunch of relevant variables in my chain output files even if they aren't being sampled over (e.g. I can get cobaya to save the value of sigma_8 for each step in the chain even if I am sampling As instead). This didn't work with a firecrown likelihood. It would be really helpful to regain this functionality and be able to sample and save a wider range of parameters with cobaya.
The text was updated successfully, but these errors were encountered:
@heatherprince In Firecrown one can use DerivedParameters (as in the des_y1_3x2pt example) to include extra information computed during the likelihood computation back in the MCMC chains. Please take a look at cobaya_evaluate.yaml:171 where we inform cobaya to include that parameter in the chains. The same appears in the cosmosis example des_y1_3x2pt.ini:20.
However, the example you gave (sigma8), that is computed by CAMB and can be included in the chains without touching Firecrown (in cobaya you simply include sigma8 in the list of derived parameters).
Usually in cobaya I can include a bunch of relevant variables in my chain output files even if they aren't being sampled over (e.g. I can get cobaya to save the value of sigma_8 for each step in the chain even if I am sampling As instead). This didn't work with a firecrown likelihood. It would be really helpful to regain this functionality and be able to sample and save a wider range of parameters with cobaya.
The text was updated successfully, but these errors were encountered: