Wrapper function for SuperLearner prediction algorithm

Share:

Description

A function in the default SuperLearner library used by tmle. This wrapper is built into SuperLearner 2.0-6, and is defined here for backwards-compatibility with older versions of SuperLearner. This function may be incorporated into any user-supplied SuperLearner library, but will not ordinarily be called directly by the user.

Usage

1
SL.glm.interaction(Y.temp, X.temp, newX.temp, family, obsWeights, ...)

Arguments

Y.temp

The outcome

X.temp

The training data set

newX.temp

The validation data set (the observations to make predictions from)

family

outcome distribution, either gaussian or binomial

obsWeights

observation weights, currently unused

...

additional arguments

Value

out

The predicted outcome on the scale of the outcome

fit

A list of elements needed for the predict method on the specific algorithm

Author(s)

Susan Gruber