CAMB sources: non-flat models

[gongbo zhao]

Dear all,

I tried to calculate the WL and count spectra for non-flat models using CAMB sources, but didn't succeed. On the website, it is said that CAMB sources doesn't support non-flat models, but I guess it is just for 21cm since there is a stop command in the 'derivs' subroutine in equations.f90,

stop 'need to add 21cm cs^2 and delta_m for non-flat'

This means that it should be able to do WL and counts for non-flat models, right? But after commenting out that stop commend in derivs, it takes ages to run for the WL spectra only. Anyone has ideas?

Thanks in advance,
Gongbo

[Antony Lewis]

It's not just 21cm, there's no non-flat support in CAMB sources at present (though probably not a major job to make corresponding changes to the non-flat code if you want to).

[gongbo zhao]

Thanks Antony. I would like to do that and publish my code here if I succeed. Could you give me some hints, like what subroutines need to be tweaked?

Thanks,
Gongbo

[gongbo zhao]

After comparing CAMB sources with CAMB, I guess the reason why CAMB sources doesn't work for the non-flat models might be the k-sampling issue. In cmbmain, there are several places saying,

if (CP%closed) then
stop 'need to check closed set k'
call SetClosedkValuesFromArr(Evolve_q%points,Evolve_q%npoints)
end if


I think that this means the k-sampling for WL, counts here needs some careful tweak, depending on the source kind and/or the median redshift?. Anybody can confirm this, or give some suggestions?

Cheers,
Gongbo

[Antony Lewis]

The non-flat bessel integration code has also not be generalized for more than 3 sources I think.