Skip to contents

Execute a visual predictive check (VPC) simulation using mrgsolve

Source: R/df_mrgsim_replicate.R
df_mrgsim_replicate.Rd

df_mrgsim_replicate() is a wrapper function for mrgsolve::mrgsim_df() that returns a data.frame containing replicates iterations of data

Usage

df_mrgsim_replicate(
  data,
  model,
  replicates,
  dv_var = "DV",
  time_var = "TIME",
  ntime_var = "NTIME",
  pred_var = "PRED",
  ipred_var = "IPRED",
  sim_dv_var = "DV",
  irep_name = "SIM",
  seed = 123456789,
  ...
)

Arguments

data

Input dataset. Must contain required variables for mrgsim_df() other than those handled by other arguments.

model

mrgsolve model object.

replicates

Number of replicates. Either an integer, or something coercible to an integer.

dv_var

Column containing the DV variable in data. Accepts bare names or strings.

time_var

Column containing the actual time variable. Accepts bare names or strings. Default is TIME.

ntime_var

Column containing the nominal time variable. Accepts bare names or strings. Default is NTIME.

pred_var

Must be "PRED" (the default). The population prediction is computed by df_mrgsim_addpred() from ipred_var with random effects zeroed, and is always returned as PRED.

ipred_var

Name of the individual prediction output captured by model (a model output, not a column of data). Returned as IPRED. Accepts bare names or strings. Default is IPRED.

sim_dv_var

Name of the simulated DV output captured by model (a model output, not a column of data). Returned as SIMDV. Accepts bare names or strings. Default is DV.

irep_name

Name of replicate variable in data. Accepts bare names or strings. Default is SIM.

seed

Random seed. Default is 123456789.

...

Additional arguments passed to mrgsolve::mrgsim_df(), including carry_out and recover to control which input columns are propagated to the output. The always-carried set (EVID, MDV, CMT, TIME, NTIME, OBSDV, and the population prediction column) is added to whatever the user passes to carry_out.

Value

A data.frame with data x replicates rows (unless obsonly=TRUE is passed to mrgsolve::mrgsim_df()) and the output variables PRED, IPRED, SIMDV, OBSDV, plus any input columns listed in carry_out / recover.

See also

Other mrgsolve wrappers: df_mrgsim_addpred(), model_mread_load()

Examples

model <- model_mread_load(model = "pkmodel")
#> Loading model from cache.
data_sad_pk <- dplyr::filter(data_sad, CMT %in% c(1,2))
simout <- df_mrgsim_replicate(data = data_sad_pk, model = model, replicates = 100,
dv_var = ODV,
carry_out = c("LLOQ", "WTBL", "FOOD"),
recover = c("USUBJID", "PART"),
irep_name = SIM)