Added functionality and tutorial to compute Poisson trispectrum from dusty sources

[ajvanengelen]

The routines in the file radio_sources.py, with tuto_radio.py loaded up a radio source model and computed the power spectrum and trispectrum by doing integrals over the counts. The trispectra are used for computing the part of the bandpower covariance matrix that is fully correlated between each band.

I generalized radio_sources.py to also work for a model of dusty sources, as will be important at high CMB frequencies. I renamed it poisson_sources.py for generality. I load up the model from Bethermin et al 2012 and sum first over redshift, and then over source flux, to get the power spectrum and trispectrum, by default at 217 GHz. I also wrote a new tuto_dusty.py file to make all the same plots as tuto_radio.py. However I did not do the part with making simple simulations (found at the bottom of tuto_radio.py)