Likelihood with |omega_K \ne 0

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Akhilesh Nautiyal(akhi)
Posts: 69
Joined: June 13 2007
Affiliation: Malaviya National Institute of Technology Jaipur

Likelihood with |omega_K \\ne 0

Post by Akhilesh Nautiyal(akhi) » June 12 2008

Hi,

I am trying to get the likelihood for the models with \Omega_k \ne 0.
I am using CAMB to get the [tex]cls[/tex] and then using these [tex]cls[/tex] to get the likelihood with wmap 5 likelihood code.

I am putting the parameters in params.ini from the parameter table given in the WMAP site.
If I am putting the parameters of the model [tex] olcdm+sz+lens[/tex] with wmpa5 data only, I am not getting a good likelihood. But if I am using the parameters for the same model with [tex]wmap5+hst[/tex], I am getting a better likelihood.
I am not getting why this is happening.
I would like to know whether the likehood code contains other data set also.

Regards,
Akhilesh

Ben Gold
Posts: 81
Joined: September 25 2004
Affiliation: University of Minnesota
Contact:

Likelihood with |omega_K \\ne 0

Post by Ben Gold » June 16 2008

The parameters in the table are the mean over the chain, not the best fit. Usually the mean is pretty close to the best fit, but in cases where the data isn't enough to constrain all the model's parameters (like OLCDM with WMAP only) there can be a bit of a difference. So it's not unexpected that the likelihood of the "OLCDM, WMAP only" mean parameters is not good. Adding in the HST data breaks a [tex]H_0 - \Omega_\Lambda[/tex] degeneracy and then the mean winds up closer to the best fit.

Post Reply