Apply prediction correction

Apply prediction correction

Usage

var_predcorr(dv_var, pred_var, lower_bound = 0)

Arguments

dv_var

Vector containing the dependent variable (DV)

pred_var

Vector containing population predictions (PRED)

lower_bound

Lower bound for prediction correction formula.

Value

A numeric vector of prediction-corrected values of dv_var

Details

The bin-median predbin is computed over the full pred_var vector passed in. Callers must group their data by a binning variable (e.g. nominal time, optionally stratified by compartment or covariates) before invoking var_predcorr() so that predbin is the median across the bin's observations, not the entire dataset.

See also

Other vectorized helpers: var_addn(), var_dosenorm()

Examples

pkmodel <- model_mread_load(model = "pkmodel")
#> Loading model from cache.
data <- df_mrgsim_addpred(
  data = dplyr::filter(data_sad, CMT != 3),
  model = pkmodel)
data <- data |>
  dplyr::filter(EVID == 0) |>
  dplyr::group_by(NTIME) |>
  dplyr::mutate(PCDV = var_predcorr(ODV, PRED)) |>
  dplyr::ungroup()