Validate MCMC, get constraints with Power Spectrum

[viajani]

Description of the issue

It is not clear which specific step is problematic in getting the constraints with peaks:

• is it the way I compute peaks? (S/N or kappa, see issue Noise estimation from real data #2)
• is it in the rescaling of the covariance? (f_sky coverage, see issue Covariance matrix for peaks #1)
• is it somewhere in the MCMC (initialisation of the walkers, prior?see notebook)

List of tasks to do within this issue:

1. show during the telecon the MCMC specifications
2. get constraints with the Power Spectrum to have a benchmark

[viajani]

Test after 18/02 telecon:
Get contours with Power Spectrum up to ell ~ 10**3