census_model_output: Compute input for the hospital census model from a filter...
In rplzzz/CovMitigation: COVID-19 infection model for Virginia

Description Usage Arguments Details Value TODO

Run the ensemble members for all counties and aggregate their results, weighted by market share.

census_model_output(
  fitdir,
  tfinal,
  scenario_name = "BAU",
  mktmin = 0.1,
  scenario = NULL,
  tstrt = NULL
)

`fitdir`	Name of directory with saved filter fit objects
`tfinal`	Max run time for the projections
`scenario_name`	Name of the scenario. Defaults to "BAU" (business as usual).
`mktmin`	Market share threshold. Counties where we have a market share below the threshold will not be run.
`scenario`	Scenario data structure returned from `Scenario`. If a data frame is passed, it will be converted with a call to `Scenario`.
`tstrt`	Start of the projection period. If omitted, the projection will start at the end of the model fit. Otherwise, the projection will start at the last historical time step prior to the requested start time. Since the model fitting is done at weekly intervals, this may not be exactly equal to `tstrt`.

If the model history is included, it will be prepended onto the projection. If not, the output will start with the projection.

The scenario structure, if present, should be a data frame that describes a schedule of hypothetical parameter changes in the future. Changes for times in the past (prior to the start of the projection) are ignored; however, the most recent such change will take effect in the first future time step. The format for the scenario structure is a data frame with the following columns:

locality: The locality the change applies to. A NA value here means the change is a default that will apply to any locality without a specified change.
time: The time at which a change in a parameter occurs. The parameter will remain at its changed value until another change is imposed.
parm: Parameter affected by the change
value: The change factor. This will be treated as a multiplier on the parameter's value at the end of the historical period. For example, a value of 1.2 will produce a 20 historical week.

Data frame suitable for input into census model. All case counts are market share weighted. Columns:

date: Simulation date
time: Simulation date expressed as days since Jan 01
scenario: Ensemble member ID number
popmkt: Market share weighted population
newCases: Daily new cases
newSympto: Daily new symptomatic cases
PopSympto: Symptomatic population
PopInfection: Infected population
fracInfection: Infection prevalence. Technically this is market share weighted, but you get the same answer whether you weight by market share or not.
PopCumulInfection: Cumulative number of infections
fracCumulInfection: Cumulative fraction of population infected. This value is the same whether or not you weight by market share