On Mon, 2002-10-28 at 19:42, Boud Roukema wrote:

What do people think of:

http://arcturus.mit.edu/cosmics/

It was the first free code for calculating density peturbations with an emphasis on CMB fluctuations.

CMBFast is based on it. But it calculates perturbations in linear regime using full set (not full, but cut anyway) of Boltzmann equations - as opposed to some radical approximation by CMBFast. The inconsistency however is only about 1-2%, but required comp. power is MUCH bigger.

In my oppinion there is no point in dealing with this soft - CMBFast is much faster and gives you practically the same accuracy. It also has more features e.g. grav. lensing, reionization.

You would need a supercomputer to run COSMICS. And just a PC to run CMBFast.

regards - michal