GetDist: H0 and sigma8 as derived parameters

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Shahbaz Alvi
Posts: 10
Joined: March 09 2020
Affiliation: The University of Ferrara
Contact:

GetDist: H0 and sigma8 as derived parameters

Post by Shahbaz Alvi » December 10 2021

I am running Planck + BAO data likelihood with the LCDM model in MontePython. I forgot to add H0 and sigma8 as derived parameters in the MontePython param file. How can I add them as a derived parameter in GetDist or is there any other way to obtain them without running the chains again?

My MCMC parameters in the param file are the following:
omega_b, omega_cdm, 100*theta_s, ln10^{10}A_s, n_s, tau_reio

Thanks.

Antony Lewis
Posts: 1812
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: GetDist: H0 and sigma8 as derived parameters

Post by Antony Lewis » December 10 2021

GetDist doesn't know anything about cosmology, so you'd have to compute the sigma8 values externally, e.g. by looping over the chain parameters and computing it.

Cobaya could do this by importance sampling original chains, but I don't know about MontePython.

Shahbaz Alvi
Posts: 10
Joined: March 09 2020
Affiliation: The University of Ferrara
Contact:

Re: GetDist: H0 and sigma8 as derived parameters

Post by Shahbaz Alvi » December 10 2021

Thanks, Anthony. I wondered if there is a way to add H0 and sigma8 as Derived Parameters in GetDist when analyzing the chains. Do you think there is a way to do this?

Antony Lewis
Posts: 1812
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: GetDist: H0 and sigma8 as derived parameters

Post by Antony Lewis » December 10 2021

Yes, but you'd need to compute signa8 in an external loop using a cosmology code. The example notebook shows hot to add derived parameters.

Shahbaz Alvi
Posts: 10
Joined: March 09 2020
Affiliation: The University of Ferrara
Contact:

Re: GetDist: H0 and sigma8 as derived parameters

Post by Shahbaz Alvi » December 10 2021

Ahhh...I see. I understand. The idea is to run the parameters in the chains (each line of each chain) with CLASS to determine sigma8 and H0. Thanks a bunch.

Post Reply