evalAugW

.evalAugW
calls TMLE to use super learner to evalute preliminary predictions for
Q(0,W) and Q(1,W) conditioning on stage 1 covariates

An inverse probability of censoring weighted (IPCW) targeted maximum likelihood estimator (TMLE) for evaluating a marginal point treatment effect from data where some variables were collected on only a subset of participants using a two-stage design (or marginal mean outcome for a single arm study). A TMLE for conditional parameters defined by a marginal structural model (MSM) is also available.

Susan Gruber

twoStageDesignTMLE

Targeted Maximum Likelihood Estimation for Two-Stage Study
Design

Mark van der Laan

evalAugW function

<dl><dt>Y</dt>
<dd>outcome vector</dd>
<dt>A</dt>
<dd>binary treatment indicator</dd>
<dt>W</dt>
<dd>covariate matrix</dd>
<dt>Delta</dt>
<dd>outcome missingness indicator</dd>
<dt>id</dt>
<dd>identifier of i.i.d. unit</dd>
<dt>family</dt>
<dd>outcome regression family</dd>
<dt>SL.library</dt>
<dd>super learner library for outcome regression modeling</dd></dl>

Arguments

.evalAugW
calls TMLE to use super learner to evalute preliminary predictions for
Q(0,W) and Q(1,W) conditioning on stage 1 covariates — evalAugW

<dl>

<dt>Y</dt>
<dd>outcome vector</dd>


<dt>A</dt>
<dd>binary treatment indicator</dd>


<dt>W</dt>
<dd>covariate matrix</dd>


<dt>Delta</dt>
<dd>outcome missingness indicator</dd>


<dt>id</dt>
<dd>identifier of i.i.d. unit</dd>


<dt>family</dt>
<dd>outcome regression family</dd>


<dt>SL.library</dt>
<dd>super learner library for outcome regression modeling</dd>

</dl>

evalAugW: .evalAugW calls TMLE to use super learner to evalute preliminary predictions for Q(0,W) and Q(1,W) conditioning on stage 1 covariates

Description

Usage

Value

Arguments