CosmoMC gaussian and HL likelihoods

Post Reply
Simone Paradiso
Posts: 3
Joined: February 03 2022
Affiliation: University of Waterloo - Waterloo Centre for Astrophysics

CosmoMC gaussian and HL likelihoods

Post by Simone Paradiso » February 03 2022

Hi all,

I am trying to understand how Cl_hat, Fiducial_Cl and covariance matrix are read from the .dataset parameter file in order to run CosmoMC with a CMB gaussian likelihood (either the fiducial gaussian and the HL, namely 'gaussian' and 'HL' in CMBlike.f90)

Does someone actually knows whether such inputs must be provided in the [lmin-lmax] multipole range, or can be given from l=2 up to some l>lmax and then CMBlikes takes care of selecting the correct rows?

Just to be clearer, I am running it in 30<l<600, only T; I have set loin and lmax accordingly in the parameter file, provided the covariance matrix as a text with (lmax-lmin+1)x(lmax-lmin+1) dimension, and given fiducial Cls in 30-600 multipole range.

Thanks to everyone who's going to help me to figure this out! :)
Simone

Antony Lewis
Posts: 1936
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: CosmoMC gaussian and HL likelihoods

Post by Antony Lewis » February 04 2022

I think the unbinned version is rather untested, so may not work, but the likelihood should take care of cutting the points/covariance as required by the inputs.

Simone Paradiso
Posts: 3
Joined: February 03 2022
Affiliation: University of Waterloo - Waterloo Centre for Astrophysics

Re: CosmoMC gaussian and HL likelihoods

Post by Simone Paradiso » February 04 2022

Thank you very much for the answer!

Simone

Post Reply