General DE with CAMB

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Niayesh Afshordi
Posts: 49
Joined: December 17 2004
Affiliation: Perimeter Institute/ University of Waterloo
Contact:

General DE with CAMB

Post by Niayesh Afshordi » June 22 2006

I want to look at a dark energy model, with a possible early dark energy component, a given w(z) and [tex]c_s \neq 1[/tex]. Is there any modification of CAMB that can do this?

Antony Lewis
Posts: 1501
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: General DE with CAMB

Post by Antony Lewis » June 23 2006

I have a version that will evolve a scalar field model, but nothing for w(z) - though this should be a fairly straightforward modification of the supplied code (just add w' term to perturbation equations - documented in the CAMB notes - and change the background).

Niayesh Afshordi
Posts: 49
Joined: December 17 2004
Affiliation: Perimeter Institute/ University of Waterloo
Contact:

Re: General DE with CAMB

Post by Niayesh Afshordi » June 23 2006

Antony Lewis wrote:I have a version that will evolve a scalar field model, but nothing for w(z) - though this should be a fairly straightforward modification of the supplied code (just add w' term to perturbation equations - documented in the CAMB notes - and change the background).
But does it work with [tex]c_s \neq 1[/tex]?

Antony Lewis
Posts: 1501
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: General DE with CAMB

Post by Antony Lewis » June 23 2006

Standard CAMB allows for constant c_s^2 \ne 1, and I think there are no complications with modifying it to become a general function of time.

Michael Doran
Posts: 41
Joined: November 22 2004
Affiliation: ITP Heidelberg
Contact:

General DE with CAMB

Post by Michael Doran » June 27 2006

In cmbeasy, I implemented a general dark energy fluid class that has arbitrary c_s^2 and w(z). However, for as long as c_s^2 remains substantially > 0, there seems to be negligble effects from the speed of sound, at least for viable dark energy models. When you approach c_s^2=0, of course, you get clustering which has a pronounced effect.
I never looked into it in detail though, i.e. I did't run a monte carlo with c_s^2 as a parameter.

Post Reply