Systematics taustats

[yomori]

This PR implements tau stats in psf_diagnostic.py

[yomori]

Cross-checked with published DES-Y3 results and get reasonable agreement, but Metacal responses have to applied manually since the Metacal implemented in TXPipe does things differently than how it was actually done in DES-Y3.

[yomori]

I remembered that there was an issue with the function load_star_type() so I fixed that as well, which was affecting how the stars were being loaded. I also modified the code such that we only compute for a specific type of star (which now gets specified in the config using star_type = 'PSF-used' or 'PSF-reserved'). The reason for this is because for certain surveys they only provide the catalog for one type of star so it will automatically crash when it tries to compute tau for the other type.