search

Chris-Pedersen / LaCE_manager

Cosmological emulator for the 1D flux power spectrum of the Lyman-alpha forest
3 stars 1 forks source link

Compare constraints from lya_theory and full_theory #38

Closed andreufont closed 2 years ago

andreufont commented 3 years ago

I wrote this on Slack in Aug 9, but before we forget I'll make it an issue on GitHub.

It would be good to compare the constraints on (Delta2_star, n_star) from a particular mock (say central or h_sim), for

Now that we have derived parameters for the latter, this should be easy to do, and an important sanity check

Chris-Pedersen commented 3 years ago

Sounds good, will run this once I've updated the linP fit procedure

Chris-Pedersen commented 3 years ago

These are running on Hypatia, will update with results in a few days

Chris-Pedersen commented 3 years ago

Forgot to mention this in our chat earlier, I wanted to show this plot below:

theory_compare

The blue chains aren't fully converged, will run them for another day or two, but so far it looks reasonable I think. Given that we have an extra degree of freedom in H_0 in the full_theory chain, we expect slightly wider constraints right? The fact there's no bias is a good sign. NB that this is also using no compression in the primordial parameters, and we get great consistency using the reconstruction from the compressed parameters. I guess this would point towards there not being a bug in the reconstruction code?

andreufont commented 2 years ago

@Chris-Pedersen - is the plot you showed last week addressing this issue?

Chris-Pedersen commented 2 years ago

theory_compare

Yes, adding the plot here for completeness!