adjust_data {simITS}

Adjust an outcome time series based on the group weights.

Description

Reweight the components of a series to match target weights for several categories. This is a good preprocessing step to adjust for time-varying covariates such as changing mix of case types.

Usage

adjust_data(
  dat,
  outcomename,
  groupname,
  Nname,
  pi_star,
  is_count = FALSE,
  include_aggregate = FALSE,
  covariates = NULL
)

Arguments

dat

Dataframe of data. Requires an N column of total cases represented in each row.

outcomename

Name of column that has the outcome to calculated adjusted values for.

groupname

Name of categorical covariate that determines the groups.

Nname

Name of column in dat that contains total cases (this is the name of the variable used to generate the weights in pi_star).

pi_star

The target weights. Each month will have its groups re-weighted to match these target weights.

is_count

Indicator of whether outcome is count data or a continuous measure (this impacts how aggregation is done).

include_aggregate

Include aggregated (unadjusted) totals in the output as well.

covariates

Covariates to be passed to aggregation (list of string variable names).

Value

Dataframe of adjusted data.

Examples

data( "meck_subgroup" )
head( meck_subgroup )
pis = calculate_group_weights( "category", Nname="n.cases", 
                               meck_subgroup, t_min=0, t_max= max( meck_subgroup$month ) )
pis

agg = aggregate_data( meck_subgroup, 
                      outcomename="pbail", groupname="category", Nname="n.cases", 
                      is_count=FALSE,
                      rich = TRUE, covariates = NULL )
head( agg )  


adjdat = adjust_data( meck_subgroup, "pbail", "category", "n.cases", pis, include_aggregate=TRUE )
head( adjdat )