Adding randomness with fake data for power spectra to make decrease cosmic variance

[Fabien Dournac]

Hello,

The issue of the cosmic variance is due to the fact that we have only one realization of universe.

Is it possible to introduce, into CAMB or CLASS, some randomness in the generation of matter/angular power spectra in order to reduce this cosmic variance ?

I think especially about the computation of quantities as [math]a_{\ell m} and [math]C_{\ell} or [math]P(k) to produce power spectra.

This way, we could have different realizations of universe since we would have multiples different fake data.

What do you think about this ?

Regards

[Fabien Dournac]

ok, Astropy library did it !