Urban Nature Access - Split radii by population groups

[phargogh]

This PR implements the final optional module in the Urban Nature Access model.

RE: #722

Checklist

• Updated HISTORY.rst and link to any relevant issue (if these changes are user-facing)
• Updated the user's guide (if needed)
• Tested the affected models' UIs (if relevant)

[dcdenu4]

Hey @phargogh, just getting around to taking a look at the testing. Just a few comments below.

• Looks like pandas.testing is not used
• I noticed a few tests for density and dichotomous kernels but not exponential, gaussian, or power. Was just wondering if all the kernels should have a test (for consistency) or if there was a reason for singling those ones out.

That's it!

[phargogh]

Thanks @dcdenu4 ! I added a couple changes to:

• Add tests for the rest of the kernel functions
• Include a link to the draft UG chapter (some of the model team were asking about it)

So, I think I've responded to everything so far! But there are still a couple things that have come out of this review so far that I think would benefit from their own PRs:

1. Implement conditional requirement at a more granular level than just the top-level keys: Implement conditional requirement for required keys anywhere in the args spec #1165
2. Define a 'safe' layer name in the target layer: Define a 'safe' layer name in Urban Nature Access #1164
3. There's still a need to define the power function's beta parameter for each possible search kernel, which means adding another column to the various tables where search radii can be defined.
4. A reviewer mentioned some more sensible naming conventions for some of the outputs, which I believe are also reflected in the paper that ultimately describes this model.

So aside from these things, I'm happy to have this merged into the feature branch if you are!

[dcdenu4]

Thanks @phargogh, I am happy to merge this into the feature branch and tackle the remaining issues in separate PRs.