Get the rxode2 model variables without compiling the model

Normalizes the input to an rxModelVars object and optionally rewrites the model text first: adding/stripping forward sensitivity equations (calcSens/calcSens2/calcSens3), adding/removing the df()/dy() Jacobian block (calcJac), dropping LHS definitions (collapseModel), or converting to an inductive linearization model (indLin).

Usage

rxGetModel(
  model,
  calcSens = NULL,
  calcJac = NULL,
  collapseModel = NULL,
  indLin = FALSE,
  calcSens2 = NULL,
  calcSens3 = NULL
)

Arguments

model: The model to get the model variables from. It can be: rxode2 model text (a character string, rxModelText, or a {} expression), a function or call whose body is the model, an rxode2 object, an rxModelVars object, an rxDll object, or anything else rxModelVars() accepts.
calcSens: boolean indicating if rxode2 will calculate the sensitivities according to the specified ODEs. May also be a character vector of the states/parameters whose first-order sensitivities (rx__sens_<state>_BY_<param>__) should be generated.
calcJac: boolean indicating if rxode2 will calculate the Jacobain according to the specified ODEs.
collapseModel: boolean indicating if rxode2 will remove all LHS variables when calculating sensitivities.
indLin: Calculate inductive linearization matrices and compile with inductive linearization support.
calcSens2: character vector (or NULL) requesting second-order sensitivities in addition to the first-order ones from calcSens. When supplied, rxode2 also generates the rx__sens_<state>_BY_<p>_BY_<q>__ compartments (the Hessian path), where p ranges over calcSens and q over calcSens2. Used, for example, for population (THETA) second-order event sensitivities. NULL (the default) skips the second-order generation.
calcSens3: character vector (or NULL) requesting third-order sensitivities in addition to the first- and second-order ones. Requires calcSens2 to also be supplied (every calcSens3 parameter needs its own already-built second-order sensitivity compartment, from the pairing of calcSens2 with calcSens3, to reference – so calcSens3 should be a subset of calcSens2, which itself should be a subset of calcSens, mirroring how calcSens2 is used everywhere else in rxode2 today). When supplied, generates the rx__sens_<state>_BY_<p>_BY_<q>_BY_<r>__ compartments, where p ranges over calcSens, q over calcSens2, and r over calcSens3, via rxExpandSens3_(). NULL (the default) skips the third-order generation.

Value

An rxModelVars object of the (possibly rewritten) model

Author

Matthew L. Fidler

Examples


rxGetModel("d/dt(depot) = -ka*depot; d/dt(central) = ka*depot - kel*central")
#> rxode2 model variables (see str to see all variables)
#> value$state: depot, central
#> value$params: ka, kel