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
@arranhamlet makes a good point that all the stan warnings might be alarming to users, and also might just be information overload that would swamp out the more meaningful diagnostic flags we have set up in model_diagnostics()
Potential solution:
suppress stan warnings by default, since wwinference() wrapper function checks for model diagnostic issues and prints a warning telling the user to run diagnostic checking if any of our defaults fail.
Curious others thoughts on this @dylanhmorris @gvegayon @seabbs
Goal
@arranhamlet makes a good point that all the stan warnings might be alarming to users, and also might just be information overload that would swamp out the more meaningful diagnostic flags we have set up in
model_diagnostics()
Potential solution:
wwinference()
wrapper function checks for model diagnostic issues and prints a warning telling the user to run diagnostic checking if any of our defaults fail.Curious others thoughts on this @dylanhmorris @gvegayon @seabbs