This function uses the zoltr R package to connect to the Zoltar database
which contains forecasts from the COVID Hub forecast project, and query it
for the specified forecast dates. It queries for all dates, computes the
proportion of dates for which the model has submitted, filters for models
that have submitted for greater than the specified proportion of forecast
dates for inclusion, and returns the vector of model names.
Usage
query_and_select_models(
prop_dates_for_incl_hub,
prop_locs_for_incl_hub,
forecast_dates,
locations,
project_name = "COVID-19 Forecasts"
)Arguments
- prop_dates_for_incl_hub
Numeric greater than 0 and less than or equal to 1 indicating the inclusion threshold for the proportion of forecast dates that a model must have submitted forecasts to be included in analysis
- prop_locs_for_incl_hub
Numeric less than 1 indicating the inclusion threshold for the proportion of the locations we expect that a model must have subbmited forecasts for to be included in analysis
- forecast_dates
vector of dates formatted in ISO8601 convention (YYYY-MM-DD) indicating the forecast dates for the analysis
- locations
vector of state abbreviations that we want to ensure the submitting teams have produced forecasts for
- project_name
name of the Zoltar project, default is
"COVID-19 Forecasts"