TMLE.updater.glm: iTMLE univariate glm learner / updater

In osofr/estimtr: Streamlined Estimation for Static, Dynamic and Stochastic Treatment Regimes in Longitudinal Data

Description

Performs a univariate TMLE update. This function is passed along as a separate learner to the SuperLearner implementation of origami package

Performs TMLE update using a main-terms GLM (logistic regression with speedglm package). This function may be passed to fit_GCOMP / fit_TMLE function. May be also used as a separate learner for iTMLE.

Usage

TMLE.updater.glm(Y, X, newX, family, obsWeights, ...)

TMLE.updater.speedglm(Y, X, newX, family, obsWeights, ...)

## S3 method for class 'TMLE.updater'
predict(object, newdata, offset, ...)

Arguments

Y	Input outcomes
X	Input design matrix with training data. Must contain a column named "offset", which contains the offsets converted to logit-linear scale.
newX	Input design matrix with test data. Same requirement as for X: must contain a column named "offset", which contains the offsets converted to logit-linear scale.
family	Link function (ignored).
obsWeights	Row-specific weights
...	Additional arguments to be passed on to origami package.
object	Results of calling TMLE.updater.speedglm or TMLE.updater.glm.
newdata	Design matrix with test data for which predictions should be obtained. May contain a column named "offset", instead of it being passed as a separate argument.
offset	Offset (on logit scale if using logistic regression update).

osofr/estimtr documentation built on Jan. 25, 2022, 8:05 a.m.