CAMB: computing logarithmic growth rate "f"

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Andreu Font-Ribera
Posts: 2
Joined: October 17 2016
Affiliation: UCL

CAMB: computing logarithmic growth rate "f"

Post by Andreu Font-Ribera » July 14 2020

Hi - I was looking for a way to compute the logarithmic growth rate "f(z)" , and Pablo Lemos pointed me to the following function in CAMB.

Code: Select all

 sigma2_vdelta_8 – array of v-delta8 correlation, so sigma2_vdelta_8/sigma_8 can define growth 
First, I think the documentation is misleading, I would consider growth to be

Code: Select all

 sigma2_vdelta_8/sigma_8**2 
Second, I couldn't figure out what "fluid" was being used here. Is this CMB+baryons (fluid=8), or does it include neutrinos (fluid=8)? I guess whenever we have massive neutrinos, "f" becomes ill-defined, and one should use f(z,k). Is there a standard / efficient way to compute f(z,k), specifying a particular fluid?

Thanks,
Andreu

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

Re: CAMB: computing logarithmic growth rate "f"

Post by Antony Lewis » July 15 2020

You can more directly use
https://camb.readthedocs.io/en/latest/results.html?highlight=get_fsigma8#camb.results.CAMBdata.get_fsigma8

(the other doc string should say "growth parameter f*sigma_8"). By default it is using config.transfer_power_var (defaults to baryons+CDM+neutrinos) and the CDM velocity. More generally you can use
https://camb.readthedocs.io/en/latest/results.html?highlight=get_sigmaR#camb.results.CAMBdata.get_sigmaR
with whatever variables you like.

However, in general it may be best to try to calculate the actual physical quantity being measured rather some rather ill-defined growth parameter.

For mode evolution you can use
https://camb.readthedocs.io/en/latest/results.html?highlight=get_redshift_evolution#camb.results.CAMBdata.get_redshift_evolution
or use the general matter power spectrum at specific requested redshifts
https://camb.readthedocs.io/en/latest/results.html?highlight=get_redshift_evolution#camb.results.CAMBdata.get_linear_matter_power_spectrum

Andreu Font-Ribera
Posts: 2
Joined: October 17 2016
Affiliation: UCL

Re: CAMB: computing logarithmic growth rate "f"

Post by Andreu Font-Ribera » July 16 2020

Thanks, I'll take a look at it and report back.

Cheers,
Andreu

Maria Berti
Posts: 1
Joined: October 26 2020
Affiliation: SISSA - Trieste

Re: CAMB: computing logarithmic growth rate "f"

Post by Maria Berti » October 29 2020

Hi!

I'd like to compute f_sigma8, but I need to use the fortran CAMB and not the python wrapper. It is a bit unclear to me how to compute it from the fortran CAMB output. Is there an already implemented way?

I would like also to have e.g. H(z) and the angular distance, as I see it's done in the demo Jupyter notebook.

Thanks!
Maria

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

Re: CAMB: computing logarithmic growth rate "f"

Post by Antony Lewis » October 29 2020

There are mostly analogous CAMBdata class methods in the fortran.
You can check the python/fortran source for the python functions to see how it's done there.

Saisandri Saini
Posts: 10
Joined: February 13 2020
Affiliation: MHRD

Re: CAMB: computing logarithmic growth rate "f"

Post by Saisandri Saini » November 08 2020

Hi
I have installed latest version of camb. I would like to modify it with Modechord. For this I have made changes to initialpower.f90,result.f90 and camb.f90 according to modified P(k) in Modechord and I have also copied the fortran files of these modules like module modpkparams, access_modpk etc., which were used in above modifications into fortran directory of CAMB.
I have added extra files (modpk, modpk_module etc.) of Modechord to the SOURCEFILE of makefile_main.
Now when I am giving make command in the fortran directory of CAMB then it is showing errors. Kindly help me.

Thanks
Saisandri

Post Reply