gaussian prior on baryon density in cosmomc

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Susana Landau
Posts: 18
Joined: December 10 2004
Affiliation: Buenos Aires University

gaussian prior on baryon density in cosmomc

Post by Susana Landau » August 27 2008

I want to run cosmomc with a gaussian prior on the baryon density from BBN. However, I can not find any information where and how to include this prior in cosmomc.

I appreciate much any help

regards

Susana

Antony Lewis
Posts: 1498
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: gaussian prior on baryon density in cosmomc

Post by Antony Lewis » August 27 2008

use_BBN=T in the input .ini file. Change the actual prior used in calc_like.f90:GetLogLikePost.

Susana Landau
Posts: 18
Joined: December 10 2004
Affiliation: Buenos Aires University

gaussian prior on baryon density in cosmomc

Post by Susana Landau » September 05 2008

Thanks Antony for your quick answer, I have still two questions

1) is it also neccesary to run getdist with the option:

adjust_priors=T

and modify the getdist.f90 ???

2) if I want to set a gaussian prior on omega_CDM, should I modify the file calc_like.f90 the subroutine GetLogLikePost in a similar way that is done for omegab?

thanks a lot

Susana

Antony Lewis
Posts: 1498
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: gaussian prior on baryon density in cosmomc

Post by Antony Lewis » September 05 2008

Whether you can get reliable results just using importance sampling (adjust priors in GetDist), depends on what effect your constraint has. Importance sampling only works well when the effect is not that large and the contraint is consistent with the other data. If importance sampling doesn't work reliably you can indeed run chains by adjusting the likelihood accordingly.

Susana Landau
Posts: 18
Joined: December 10 2004
Affiliation: Buenos Aires University

gaussian prior on baryon density in cosmomc

Post by Susana Landau » September 06 2008

thanks Antony again,

I still have another doubt, when running cosmomc with the gaussian prior on omegab, i first tried the standard value in cosmomc omegabh2=0.022 \pm 0.002. However, when plotting the chains, I see that omegabh2 takes values between 0.023 and 0.025, which are not those indicated by the prior. ,
Am I doing something wrong while running cosmomc? Should omegabh2 take only values between 0.0020 and 0.0024?

thanks a lot again for your quick answers

[/img]

Post Reply