Create mock submission scores

For each forecast date, we want to create a mock submission to the Hub so forecasts for all states + territories for each scenario. Since we know the scores are independent and we just do a weighted average over them to get an overall score, we can just take the scores directly from the "no_wastewater" scenario whenever a wastewater score is missing, add it to the dataframe, and continue to concatenate until we get a full score dataframe with all combinations of forecast dates, locations, and scenarios, labeled according to scenario. We will still know which ones are from which model with the `model“ column

Usage

create_mock_submission_scores(
  all_scores,
  name_of_replacement_model = "no_wastewater"
)

Arguments

all_scores

This a dataframe of scores for all the successful model runs. Because there are locations in each scenario where the wastewater data is missing, this is not all the combinations of scenarios, forecast dates, and locations

name_of_replacement_model

the the string identifier in the scenario of all_scores that should be used to replace the missing locations for each scenario and forecast date. Default is no_wastewater as we will typically use the hospital admissions only model for this.

Value

all_submission_scores which will be an expanded dataframe with scores for each day with evaluation data for all combinations of forecast dates, locations, and scenarios, with the name of replacement model scores filling in when scores are missing.