Page 1 of 1

CAMB - arbitrary power spectra

Posted: August 08 2006
by Brian Powell
Hello -

I am currently trying to modify CAMB to take in arbitrary power spectra, however, I am unsure of the best way to sample the k-space. A naive logarithmic sampling across the full range of k values causes CAMB to return incorrect CMB spectra. Also, given a prescription for k sampling, can one compute the dk vector by simply taking the difference between subsequent k values?
As I am very new to the software, any help would be greatly appreciated.

Re: CAMB - arbitrary power spectra

Posted: August 08 2006
by Antony Lewis
You should only have to modify PowerTilt.f90 - as long as the values returned by your ScalarPower are fairly smooth as a function of k CAMB should not have any problem.