Model library

[mileslucas]

It would be nice to start implementing models which have analytical evidence formulas, both to show in examples in the docs, and to use as unit tests. I think we can create a new submodule, Models, and write a bunch of log-likelihood functions and priors, using const GMM = NestedModel(log_GMM, prior_GMM). Ideally these models can be written without any additional dependencies than are already present.

For inspiration, here are some papers/repos which have models defined, usually in python

• https://arxiv.org/pdf/2012.15286.pdf
• https://dynesty.readthedocs.io/en/latest/examples.html#
• https://arxiv.org/pdf/2101.09675.pdf (sec. 6)