use cosmology API instead of deprecated cosmology package

[ntessore]

Switch to use the cosmology API instead of the old, deprecated cosmology package.

There is a work-in-progress CAMB wrapper, which needs to be checked for feature-completeness against all uses of cosmo in GLASS. I don't think we use anything that's not covered by the cosmology API currently (growth functions, power spectra).

Since the cosmology API functions have no dimensionless equivalent to the cosmology package, there is probably some amount of testing needed in e.g. the lensing module.

cc @nstarman

[nstarman]

How can I help?

[ntessore]

How can I help?

I think this really only needs some time to sit down and make the changes -- I just wanted to keep you in the loop on this issue.

(And it will probably take longer to write the docs on the cosmology object than to make code changes, there's surprisingly little cosmology use in our library.)

[paddyroddy]

I've made a start on this in #397. My questions/thoughts:

1. What is the equivalent of cosmology.Cosmology.dc_inv? We have cosmology.api.StandardCosmology.comoving_distance, but I need the inverse of that
2. NLA model for intrinsic alignments #21 refers to
   • cosmology.Cosmology.h
   • cosmology.Cosmology.Om
   • cosmology.Cosmology.gf
     None of which I believe exist in. I have reverted this in gh-124: use cosmology.api over the cosmology package #397.
3. What is the equivalent of cosmology.Cosmology.from_camb(pars)? I'm not sure if it's just not ready yet

[paddyroddy]

Discussed in a meeting.

• Looks like NLA model for intrinsic alignments #21 was using an unreleased version of the old https://github.com/glass-dev/cosmology package. Have raised gh-399: hide kappa_ia_nla #400 to tackle this.
• Old package patched missing things like Hubble parameter, discussed that the new API does not do this
• In practice when using the cosmology.api one would make their own CAMB instance and pass it to the API