CosmoMC: Priors

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Mario Acero
Posts: 28
Joined: November 15 2006
Affiliation: Universidad del Atlantico
Contact:

CosmoMC: Priors

Post by Mario Acero » December 20 2006

Hi everybody,

I have changed one of the derived quantities ([tex]r_{10}[/tex]) in the GetDist.f90 file, writing, in the MapParameters subroutine the following:

Code: Select all

invars(2+20)=(931400*invars(2+6)*invars(2+2))/(invars(2+21)*invars(2+21))


It is to define the neutrino mass as
[tex]
m_{\nu} = 93.14*f_{\nu}*\omega_{dm} (eV).
[/tex]


The results I got are sowing a double-peaked distribution for the neutrino mass [tex]m_{\nu}[/tex] and the fraction of DM in form of neutrinos [tex]f_{\nu}[/tex] (figure), and I think it should not be like that.
Someone suggested that there could be a problem with the priors, but I am no sure about that.

Do you have eny suggestion about this?

Thanks a lot,

MarioA.

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

CosmoMC: Priors

Post by Ben Gold » December 21 2006

It doesn't surprise me much: a heavy enough neutrino is going to be non-relativistic during the radiation era and indistinguishable from cdm! I'm more curious about why the likelihood actually seems to start going back down again for m_\nu > 8 eV or so.

Anze Slosar
Posts: 183
Joined: September 24 2004
Affiliation: Brookhaven National Laboratory
Contact:

Re: CosmoMC: Priors

Post by Anze Slosar » December 21 2006

Ben Gold wrote:It doesn't surprise me much: a heavy enough neutrino is going to be non-relativistic during the radiation era and indistinguishable from cdm! I'm more curious about why the likelihood actually seems to start going back down again for m_\nu > 8 eV or so.
But I think that it has to be quite a bit heavier than that to reach cdm limit, of the order of keV or so, right?

Mario, I am a bit confused about your equation, param 2 is \omega_{dm}, so there is no need to divide by H_0^2 (is this what are you doing?)? Also, I don't think a double peaked f_\nu is right to start with...

Mario Acero
Posts: 28
Joined: November 15 2006
Affiliation: Universidad del Atlantico
Contact:

CosmoMC: Priors

Post by Mario Acero » December 21 2006

Anze,

Thanks for your comment. You are write, that is what I was doing. I was confused with the definitions.

On the other hand, I do not know yet the possible error I am making here. The point is that, if I do not consider the neutrino mass, then the [tex]f_{\nu}[/tex] distribution is not double peaked, so...

What do you think?

MarioA.

Post Reply