Setup Pred function based on RxODE object.

Source: R/rxJacobian.R

This is for the so-called inner problem.

rxSEinner(
  obj,
  predfn,
  pkpars = NULL,
  errfn = NULL,
  init = NULL,
  grad = FALSE,
  sum.prod = FALSE,
  pred.minus.dv = TRUE,
  only.numeric = FALSE,
  optExpression = TRUE,
  interaction = TRUE,
  ...,
  promoteLinSens = TRUE,
  theta = FALSE,
  addProp = c("combined2", "combined1")
)

rxSymPySetupPred(
  obj,
  predfn,
  pkpars = NULL,
  errfn = NULL,
  init = NULL,
  grad = FALSE,
  sum.prod = FALSE,
  pred.minus.dv = TRUE,
  only.numeric = FALSE,
  optExpression = TRUE,
  interaction = TRUE,
  ...,
  promoteLinSens = TRUE,
  theta = FALSE,
  addProp = c("combined2", "combined1")
)

Arguments

obj

RxODE object
predfn

Prediction function
pkpars

Pk Pars function
errfn

Error function
init

Initialization parameters for scaling.
grad

Boolaen indicated if the the equations for the gradient be calculated
sum.prod

A boolean determining if RxODE should use more numerically stable sums/products.
pred.minus.dv

Boolean stating if the FOCEi objective function is based on PRED-DV (like NONMEM). Default TRUE.
only.numeric

Instead of setting up the sensitivities for the inner problem, modify the RxODE to use numeric differentiation for the numeric inner problem only.
optExpression

Optimize the model text for computer evaluation.
interaction

Boolean to determine if dR^2/deta is calculated for FOCEi (not needed for FOCE)
promoteLinSens

Promote solved linear compartment systems to sensitivity-based solutions.
theta

Calculate THETA derivatives instead of ETA derivatives. By default FALSE
addProp

one of "combined1" and "combined2"; These are the two forms of additive+proportional errors supported by monolix/nonmem:

combined1: transform(y)=transform(f)+(a+b*f^c)*eps

combined2: transform(y)=transform(f)+(a^2+b^2*f^(2c))*eps

Value

RxODE object expanded with predfn and with calculated sensitivities.

Author

Matthew L. Fidler