CDCgov / ww-inference-model

An in-development R package and a Bayesian hierarchical model jointly fitting multiple "local" wastewater data streams and "global" case count data to produce nowcasts and forecasts of both observations
https://cdcgov.github.io/ww-inference-model/
Apache License 2.0
17 stars 2 forks source link

Come up with a way to handle if `sum(site_pop) > total_pop` #136

Open kaitejohnson opened 2 months ago

kaitejohnson commented 2 months ago

Goal

In general, this shouldn't be the case and we should make sure to at least throw a warning to users because its probably something wrong with their input data. However, we also probably want something to work around this if this is the case, since we did see this in production pipeline last season. We just don't want it to happen silently.

Requirements

seabbs commented 2 months ago

Can you try and roughly do some estimation of cross catchment contamination?

kaitejohnson commented 2 months ago

I'm not sure -- I think you could use county fip codes potentially, but the reality is that a wastewater catchment area could be touching multiple counties and not be overlapping.

The one instance that we see this is in a very small jurisdiction -- my guess is that the wwtp serves people in other states so its actually not overlapping just extending beyond the jurisdictional pop.

I think for a package though, we mostly want to make sure this is flagged bc further investigation would need to happen outside of the package