Support for pstokes in the pol_pairs arg of pspec_run if we use an efield beam.

[aewallwi]

If we include pstokes in the pol_pairs arg of pspec_run (e.g. [(pI, pI)]) we get an error because pspecbeam.PSpecBeamUV uses efield_to_power rather then efield_to_pstokes when initializing its uvbeam object so it does not contain any information on the pstokes beams.

hera_pspec/hera_pspec/pspecbeam.py

Line 432 in 10c290f

self.primary_beam.efield_to_power(inplace=True)

Using efield_to_pstokes will not support linear pol power spectra though (maybe I'm wrong about this?) so I propose including a copy of self.primary_beam in pspecbeam.PSPecBeamUV called self.pstokes_primary_beam that stores the pstokes pols and gets referred to when computing Omegap and Omegapp for pstokes pols.

[nkern]

You should be using a pI power beam to normalize the power spectra. and if you want to use X or Y power beam for power spectra, you should be able to get it with existing infrastructure between pyuvdata and hera_pspec without having to make a copy of the beam itself. whether pspec_run knows how to do this is another thing, I think, but in principle the tools should exist