TR: Individualized treatment regime based on PAL or SAS.

In ITRSelect: Variable Selection for Optimal Individualized Dynamic Treatment Regime

Description

Recommend individualized treatment regime for future patients, based on the penalized A-learning method, or sequential advantage selection method.

Usage

TR(object, x1, a1 = NULL, x2 = NULL, stage = 1)

Arguments

object	Fitted object of class "PAL" or "SAS".
x1	A matrix consisting of future patients baseline covariates.
a1	A vector consisting of future patients first treatments. Not needed if stage = 2.
x2	A matrix consisting of future patients intermediate covariates. Not needed if stage = 2.
stage	Outputs the first-stage decision rule for future patients if stage = 1. Otherwise, outputs the second-stage decision rule for future patients.

Value

A vector of individualized treatments tailored for future patients.

Author(s)

Chengchun Shi

See Also

PAL, SAS

Examples

## load simulated STARD data
data(SSTARD.twostage)
## estimate individualized treatment regime using SAS
result <- SAS(Y~X1|A1|X2|A2, data=SSTARD.twostage)
## make the recommendation 
TR(result, x1=SSTARD.twostage$X1, a1=SSTARD.twostage$A1, x2=SSTARD.twostage$X2, stage=2)

ITRSelect documentation built on May 1, 2019, 10:56 p.m.