Compute the posterior of the difference between model predictions.

Usage

compute_forecast_differences(
  forecast_date,
  location,
  scenario,
  raw_output_dir,
  offset = 1
)

Arguments

forecast_date: forecast_date for which to fits recent data trends. Data for an actual forecast must already have been produced and exist in raw_output_dir.
location: Location for which to fit trends.
scenario: Wastewater data availability scenario to analyze.
raw_output_dir: Directory containing raw output .rds files. Used to obtain the admissions and wastewater data used in fitting the forecasting model.
offset: to add to incidences when computing a log difference. Default 1.

Value

Posterior of the difference in total incidence across the full nowcast/forecast period. Saves the incidence differences by day and the total incidence difference to disk as a side effect.