CosmoMC: Best-fit power spectrum?

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Hans Kristian Eriksen
Posts: 60
Joined: September 25 2004
Affiliation: ITA, University of Oslo

CosmoMC: Best-fit power spectrum?

Post by Hans Kristian Eriksen » April 04 2005

I'm feeling a little stupid here now, because there must surely be some easy way to do this... Anyway, what is the easiest way of getting the best-fit C_l power spectrum from CosmoMC, after a run has been completed?

I understand that this information should be stored in some .data files, but unfortunately, they are only optionally outputted (and I don't even find which option that toggles it right now...), and therefore not of much use for already computed chains.


Antony Lewis
Posts: 1646
Joined: September 23 2004
Affiliation: University of Sussex

Re: CosmoMC: Best-fit power spectrum?

Post by Antony Lewis » April 06 2005

There isn't actually any automated way to do this, sorry. (feel free to write one...)

After running GetDist, the .likestats file has the parameters of the best fit sample. You can feed these manually into CAMB's input file to generate a spectrum (obviously you need the derived parameter H_0 rather than \theta for CAMB).

Change the indep_sample parameter in the the cosmomc input file to make it produce .data files for later importance sampling.

Post Reply