Example model output data representing mean point forecasts
example_mean_model_output.RdThese are based on a subset of forecasts of daily incident hospital
admissions for COVID-19 that were submitted to the US COVID-19 Forecast hub,
which have been lightly modified for expository purposes. To obtain point
forecasts, we used the distfromq package for R to estimate the quantile
function, drew 100,000 samples from the distribution using inverse transform
sampling, and computed the mean of those samples.
Format
example_mean_model_output
A data frame with 16 rows and 8 columns:
- origin_date
Date: Reference date for the forecast.
- horizon
integer: Forecast horizon, in days.
- location
character: FIPS codes identifying the location being forecasted.
- target
character: The forecast target. All values are "inc covid hosp".
- output_type
character: Output type used to represent the forecast. All values are "mean".
- output_type_id
NA: This column is not used when the output type is "mean". All values are "NA".
- value
numeric: The mean of the forecast distribution.
- model_id
character: Identifier for the model that produced the forecast.