Wastewater Model

[sbidari]

This PR replicates the WWW model as implemented here.

---------------fixed-----------------------------------------------
There is one issue with using predictive methods, specifically for the observed hospital admissions. Currently, the posterior predictive distribution for observed_hospital_admissions does not extend beyond the available datapoints because of how it is defined in the model. This was done to account for possible differences in the length of reporting time for wastewater and hospital admissions data, following the approach in the stan code. I am not sure how to circumvent this issue for the pyrenew model and would love to hear ideas/suggestions.
@dylanhmorris @damonbayer @kaitejohnson

[damonbayer]

@sbidari Can you share the rendered qmd file somewhere for me to look at?

[damonbayer]

@sbidari I think we should also include an equivalent of demos/hosp_only_ww_model/model_comp.qmd so we can easily compare results to the stan model.

[sbidari]

@sbidari Can you share the rendered qmd file somewhere for me to look at?

Rendered qmd files here https://cdc.sharepoint.com/:f:/r/teams/CenterforForecastingandOutbreakAnalytics/Shared%20Documents/General/02%20-%20Predict/Real%20Time%20Monitoring%20(RTM)%20Branch/Short%20Term%20Forecasts/PyRenew%20Wasewater%20Model?csf=1&web=1&e=3Y4JuM

[damonbayer]

Given that this runs and produces plausible results, I am going to merge to main now. I will open an issue for further improvements and evaluation to be done next year.