df_pcdv
is a helper function to perform prediction-correction
of observed or simulated depedent variables.
df_pcdv(
data,
bin_var = "NTIME",
strat_vars = NULL,
dvpred_vars = c(PRED = "PRED", DV = "DV"),
lower_bound = 0
)
Input dataset
Exact binning variable. Default is "NTIME"
.
Stratifying variables. Default is NULL
.
Names of variables for the dependent variable and population model prediction. Must be named character vector.
Defaults are "PRED"
and "DV"
.
Lower bound of the dependent variable for prediction correction. Default is 0
.
A data.frame containing one row per unique combination of
bin_var
and strat_vars
and new variable PCDV
containing
prediction-corrected observations.
model <- model_mread_load(model = "model")
#> Loading model from cache.
data <- df_addpred(data_sad, model)
simout <- df_pcdv(data, dvpred_vars = c(DV = "ODV", PRED = "PRED"))