Write ensemble forecast summary to disk. — write_ref_date_summary

This function generates and writes ensemble-only forecast data.

Usage

write_ref_date_summary_ens(
  reference_date,
  base_hub_path,
  hub_reports_path,
  disease,
  horizons_to_include = c(0, 1, 2),
  population_data = hubhelpr::population_data,
  excluded_locations = NULL,
  output_format = "csv",
  targets = NULL,
  n_models_for_reporting = 2,
  overwrite_existing = FALSE
)

Arguments

reference_date: character, the reference date for the forecast in YYYY-MM-DD format (ISO-8601).
base_hub_path: character, path to the forecast hub directory.
hub_reports_path: character, path to forecast hub reports directory.
disease: character, disease name ("covid" or "rsv").
horizons_to_include: integer vector, horizons to include in the output. Default: c(0, 1, 2).
population_data: data frame with columns "location" and "population". Default: population_data.
excluded_locations: character vector or named list specifying US state abbreviations to exclude. If a character vector, locations are excluded across all targets. If a named list, names should be target names (or "all" for global exclusions) mapping to character vectors of abbreviations. Default: NULL.
output_format: character, output file format. One of "csv", "tsv", or "parquet". Default: "csv".
targets: character vector, target name(s) to filter forecasts. If NULL (default), does not filter by target.
n_models_for_reporting: integer, minimum number of designated model submissions required to include an ensemble forecast in the report. Default: 2.
overwrite_existing: logical. If TRUE, overwrite existing files. Default: FALSE.

Value

invisibly returns the file path where data was written