Create a new stan model using a template

Usage

new_stan_model(
  parameters,
  parameter_definitions,
  fixed = NULL,
  variable_definitions = NULL,
  ode = NULL,
  covariate_definitions = NULL,
  solver = c("pmx_solve_onecpt", "pmx_solve_twocpt", "pmx_solve_rk45", "pmx_solve_adams",
    "pmx_solve_bdf"),
  template = "template.stan",
  obs_types = NULL,
  obs_cmt = NULL,
  scale = NULL,
  custom_ipred = NULL,
  verbose = FALSE
)

Arguments

parameters: list of parameter values, e.g. list(CL = 5, V = 50).
parameter_definitions: list of parameter definitions. See example below.
fixed: optional, vector of parameters to fix and not sample.
variable_definitions: for models using the analytical solvers, extra variable definitions evaluated at each step through the event table. See busulfan example in vignettes for example.
ode: code for differential equation system, if not the analytic solver is used.
covariate_definitions: list of covariate definitions. See example below.
solver: Torsten solver to use. Starts with pmx_solve_.
template: internal Stan template file to use. Defaults to template.stan
obs_types: vector of observed data types. If NULL, will use "pk".
obs_cmt: observation compartment, or list of observation compartments for each observation type when multiple observation types. E.g. 2 or list(pk = 2, pd = 8).
scale: scale definition. Can be either parameter expression or number.
custom_ipred: list of custom definitions for individual predictions for each observation type. Sometimes this is needed when more flexibility is needed than the scale parameter provides.
verbose: verbosity

Value

filename

Examples


parameters <- list(CL = 5, V1 = 50, Q = 5, V2 = 100)
iiv <- list(CL = 0.27, Q = 0.49, V1 = 0.15, V2 = 1.3)
ruv <- list(add = 1.6, prop = 0.15)

model_file <- new_stan_model(
  parameters = parameters,
  parameter_definitions = list(
   "CL" = "CL * (1.0 + 0.0154 * ((CRCL[j] * 16.6667) - 66.0))",
   "Q"  = "Q",
   "V1" = "V1 * WT[j]",
   "V2" = "V2 * WT[j]",
   "KA" = "0" 
  ),
  covariate_definitions = list(
    "CRCL" = "real", # CrCl in L/hr
    "WT" = "real"    # WT in kg
  ),
  solver = "pmx_solve_twocpt",
  obs_cmt = 2,
  scale = "(V1 * mean(WT))",
  verbose = TRUE
)
#> assuming scale definition (V1 * mean(WT)) for all observation types
#> assuming observation type 2 for all observation types