Sort out mass definitions - mean or critical density?

[jhod0]

There are two ways of defining a halo of mass M - either Delta times the mean density of the universe or the critical density of the universe. These should not be confused. A few things to do:

• An optional critical-mean mass definition converter exists in tests/cosmology.py at commit 13d4a66. Making this standard would add a dependency on the Colossus library.
• The Battaglia profile is calibrated for the critical mass definition. Make sure to clarify this in the documentation.
• The current two-halo term confuses these - it assumes the halo mass function and bias use the same mass definition as the profile, but a commonly used mass function is the Tinker function which uses mean density. It's almost intentionally confusing... in any case, allow either but document it clearly.

[jhod0]

Fixed by #5, which takes a user-given table for the M_{\Delta m} <-> M_{\Delta c} conversions