Run an MIPD trial — run_trial • mipdtrial

This function is a wrapper function to make it easier to run a basic MIPD trial with minimal function calls. For fine-tune control over logic, it may be easier to compose each step of the precision-dosing workflow yourself. For an example, see the "Busulfan MIPD" vignette.

Usage

run_trial(
  data,
  design,
  cov_mapping,
  n_ids = NULL,
  seed = 0,
  verbose = FALSE,
  progress = TRUE,
  threads = 1
)

Arguments

data: data set to use for simulation, a data frame. Should contain covariates in columns, corresponding to values in cov_mapping and a column for ID, as a patient identifier.
design: trial design, created using create_trial_design()
cov_mapping: a named vector indicating correspondence between model covariates and data frame columns. Vector names should reflect all covariates defined by the model, and vector values should represent the associated column in the provided data set.
n_ids: number of subjects to use in simulated trial. If not specified, will use all subjects in data.
seed: argument passed to set.seed() to set the seed for all random variability generation to ensure reproducibility of results. Even when run in parallel, reproducibility is ensures since each thread will receive a reproducible seed number.
verbose: verbose output?
progress: show progress bar? (default is TRUE)
threads: number of threads to run the simulations on. By default the simulations will run on 4 cores, or less if less are available (one core will always be reserved). The user can override by specifying manually.

Details

This function performs the following steps:

Dose the patient using the function specified in initial_dose_method.
Collect levels at the specified times (tdm_times).
Adjust dosing at each specified dose (adjust_doses_at) using the logic specified in dose_optimization_method
Return an object with data collected during and at the end of the MIPD trial.