Compiled nlmer sensitivity model (prediction + gradients)

Returns a list with:

• thetaGrad: compiled rxode2 model computing rx_pred_ and rx__sens_rx_pred__BY_THETA_i___ for each estimated parameter. When eventSens = "jump" this model carries rxode2's analytic event-jump sensitivities (loaded/activated by nlmixr2est::.nlmSetupEnv()).

• predOnly: compiled rxode2 model for prediction only

• eventTheta: integer vector flagging THETA[i] that need Shi2021 finite differences for dosing parameters (all 0 under eventSens = "jump")

• paramNames: character vector of parameter names in THETA[i] order

Usage

# S3 method for class 'nlmerSensModel'
rxUiGet(x, ...)

Arguments

x

rxUiGet list(ui)

...

additional arguments (currently ignored)

Details

The list follows the modelInfo contract of nlmixr2est::.nlmSetupEnv() so the nlm C machinery can load it once and keep it resident.