computation via alloscore and distfromq

[aaronger]

Let's consolidate issues #13 and #14 here. The questions are whether, where, and how to explain the computation of allocation scores from hub hosp quantile forecasts.

• whether: seems necessary. I'd rather not just say "it's all in the code repos" even if reviewers let it slide.
• where: seems like after \paragraph{Step 3: Define the scoring rule.} and/or in \subsubsection{Allocation Score} would be the natural insertion points for the current draft. But maybe this needs a new section?
• how: definitely only a sketch in the text. Some more detail in the supplement. Wondering though if an iteration picture like the one in my slides would be appropriate or useful.

[elray1]

I agree that we should do this. r.e. where, in part I think that the distfromq piece is just an add-on that is specific to our application. We don't talk about the fact that our application is working with quantile forecasts until the application section. So I might add the discussion about distfromq somewhere after that?

[aaronger]

And then maybe after the distfromq distributions are described we can talk about using them to iterate through allocations with alloscore...

[elray1]

just noting that currently the only discussion of this sits in the supplement but it seems important to mention this at some level in the main text too

[aaronger]

just noting that currently the only discussion of this sits in the supplement but it seems important to mention this at some level in the main text too

Maybe will be easier to do this once it starts to solidify in the supplement. (What I'm working on right now...)

[aaronger]

Well, stuff is there now... not sure how much easier it makes mentioning this in the main text.