Added a new metric (energy score)

[thanasibakis]

What's new:

• A new metric: energy score, which is a multivariate generalization of CRPS
• proportions_mae is now proportions_mean_norm, to be a bit more correct in our naming
  • Correct me if I'm misremembering, but "mean absolute error" is defined on scalars, not vectors, and we were summing the MAE over lineages to resolve this. This is equivalent to taking the "mean L1 norm of the error", so I use that name now
• I realized quickly that joining samples and data was going to happen in every metric function, so that is its own function now

[thanasibakis]

I took a stab at what unit tests could look like for this, as @swo mentioned...

It's not perfect, but if we choose to use the L1 norm everywhere, then these expectations have analytic solutions for the case where the forecast distribution is normal. What I did was generate lineage counts per division-day, compute true lineage proportions per division-day, and then created 1000 "forecasts" of this by sampling from a normal centered around the true proportion.

These clearly aren't valid proportions, as they don't sum to 1 and could even be negative, but the L1 norm and energy score computations don't consider these constraints at all and are still valid. I was able to compare the estimates to what the closed-form solutions dictate they should be, and they match up to decent tolerance

[thanasibakis]

Added the hand-derived unit tests for fixed dataframes as discussed! @swo @afmagee42

pytest passes in the current state of things

[thanasibakis]

(Just rebased on the mkdocs commit)