simulate.SaemixObject: Perform simulations under the model for an saemixObject object

Description

This function is used to simulate data under the empirical design, using the model and estimated parameters from a fit.

Usage

# S3 method for SaemixObject
simulate(
  object,
  nsim,
  seed,
  predictions,
  outcome = "continuous",
  res.var = TRUE,
  uncertainty = FALSE,
  ...
)

Arguments

object: an saemixObject object returned by the saemix function
nsim: Number of simulations to perform. Defaults to the nb.sim element in options
seed: if non-null, seed used to initiate the random number generator (defaults to NULL)
predictions: Whether the simulated parameters should be used to compute predictions. Defaults to TRUE for continuous data, and to FALSE for non-Gaussian data models. If FALSE, only individual parameters are simulated.
outcome: the type of outcome (used to specify TTE or RTTE models)
res.var: Whether residual variability should be added to the predictions. Defaults to TRUE
uncertainty: Uses uncertainty (currently not implemented). Defaults to FALSE
...: additional arguments, unused (included for compatibility with the generic)

Author

Emmanuelle Comets emmanuelle.comets@inserm.fr, Audrey Lavenu, Marc Lavielle.

Details

The simulated data can then be used to produce Visual Predictive Check graphs, as well as to compute the normalised prediction distribution errors (npde).

This function replaces the previous function (simul.saemix), which will be deprecated in future versions but can still be called as previously for compatibility purposes.

References

Brendel, K, Comets, E, Laffont, C, Laveille, C, Mentre, F. Metrics for external model evaluation with an application to the population pharmacokinetics of gliclazide, Pharmaceutical Research 23 (2006), 2036-2049.