Redshift-dependent residual in the NLA model

[jharno]

Looking at the Compare_Mocks_theory notebook, there is a small residual mismatch between mocks and theory. I want to track this down and improve the match. I first need to make sure that the IA model really is the redshift independent NLA, without the [(z+1)/(z_0+1)]^(eta) term. Once that's verified, and that nothing else has become apparent, I might need to come up with an empirical calibration.