Most if not all of the arguments to `cmdstanr` `sample()` should be optionally configurable from the `wwinference()` interface

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

16 stars 2 forks source link

Most if not all of the arguments to `cmdstanr` `sample()` should be optionally configurable from the `wwinference()` interface #174

Closed dylanhmorris closed 2 weeks ago

dylanhmorris commented 4 weeks ago

Right now many are not. A power user might wish to configure them.

athowes commented 3 weeks ago

What about a ... version here? Like

wwinference <- function(ww_data,
                        count_data,
                        forecast_date = NULL,
                        calibration_time = 90,
                        forecast_horizon = 28,
                        model_spec = get_model_spec(),
                        generate_initial_values = TRUE,
                        initial_values_seed = NULL,
                        compiled_model = compile_model(),
                        ...)

And then passed through to Stan?