Format forecasts for submission — format_for

This function prepares forecasts to adhere to probabilistic forecast submission guidelines for consortia such as FluSight.

Usage

format_for_submission(
  .forecasts,
  method = "ts",
  .target = "wk ahead inc flu hosp",
  format = "hubverse",
  horizon_shift = 1
)

Arguments

.forecasts: Forecasts to be formatted for submission; if method is "ts" this should be forecasts from ts_fit_forecast; otherwise this must be a tibble with forecast output (e.g., output from glm_forecast) with a column designating "location"
method: Method for forecasting; default is "ts" which will trigger the use of ts_format_for_submission internally
.target: Name of the target in the forecast; default is "wk ahead inc flu hosp"
format: The submission format to be used; must be one of "hubverse" or "legacy"; default is "hubverse"
horizon_shift: Number of horizons to shift backwards to align with reference date; only used if format is "hubverse"; default is 1

Value

A named list of tibbles with probabilistic forecasts (one for each model), formatted for submission.

If format is "hubverse" each tibble will have the following columns:

reference_date: Date of reference for forecast submission
horizon: Horizon for the given forecast
target: Name of forecasted target
target_end_date: Last date of the forecasted target (e.g., Saturday of the given epidemiological week)
location: Name or geographic identifier (e.g., FIPS code) for location for the given forecast
output_type: Type of forecasted value (e.g., "quantile")
output_type_id: The quantile for the forecasted value if output_type is "quantile"
value: The forecasted value

If format is "legacy" each tibble will have the following columns:

forecast_date: Date of forecast
target: Horizon and name of forecasted target
target_end_date: Last date of the forecasted target (e.g., Saturday of the given epidemiological week)
location: Name or geographic identifier (e.g., FIPS code) for location for the given forecast
type: One of either "point" or "quantile" for the forecasted value
quantile: The quantile for the forecasted value; NA if "type" is "point"
value: The forecasted value

References

https://github.com/signaturescience/FluSight-forecast-hub/tree/main/model-output#forecast-file-format

https://github.com/cdcepi/Flusight-forecast-data/blob/master/data-forecasts/README.md

Examples

if (FALSE) { # \dontrun{
# Get raw data from healthdata.gov
h_raw <- get_hdgov_hosp(limitcols=TRUE)

# Prep, and make a tsibble
prepped_hosp <- prep_hdgov_hosp(h_raw, statesonly=TRUE)
prepped_hosp_tsibble <- make_tsibble(prepped_hosp,
                                     epiyear = epiyear,
                                     epiweek=epiweek,
                                     key=location)
# Limit to only Virginia and US
prepped_hosp_tsibble <-
  prepped_hosp_tsibble %>%
  dplyr::filter(location %in% c("US", "51"))

# Fit a model
hosp_fitfor <- ts_fit_forecast(prepped_hosp_tsibble,
                               horizon=4L,
                               outcome="flu.admits",
                               covariates=TRUE)

# Format for submission
formatted_list <- format_for_submission(hosp_fitfor$tsfor, method = "ts", format = "legacy")
formatted_list
} # }