Skip to contents

Create a simulator function from a fitted Single-index outcome model

Source: R/simulator_factory.R
make_single_index_simulator.Rd

This factory builds a closure that can be used to simulate outcome values from a fitted SensIAT::Single-index-outcome-model. The returned function samples from the estimated conditional distribution of the outcome given covariates using the Nadaraya-Watson estimator implemented in pcoriaccel_NW.

Usage

make_single_index_simulator(outcome_model, covariate_mapping = NULL)

Arguments

outcome_model

A fitted object of class SensIAT::Single-index-outcome-model.

covariate_mapping

Optional named character vector mapping expected covariate names (e.g. prev_outcome, time, delta_time) to the names used in the original model formula. If NULL, the factory will attempt to use prev_outcome, time, and delta_time directly.

Value

A function with signature function(prev_outcome, time, delta_time, newdata = NULL) which returns a sampled outcome value consistent with the fitted model.