subsemble (version 0.1.0)

predict.subsemble: Predict method for a 'subsemble' object.

Description

Obtains predictions on a new data set from a subsemble fit. May require the original data, X, if one of the learner algorithms uses the original data in its predict method.

Usage

# S3 method for subsemble
predict(object, newx, x = NULL, y = NULL, ...)

Arguments

object

An object of class 'subsemble', which is returned from the subsemble function.

newx

The predictor variables for a new (testing) data set. The structure should match x.

x

Original data set used to fit object.

y

Original outcome used to fit object.

…

Additional arguments passed to the predict.SL.* functions.

Value

pred

Predicted values from subsemble fit.

subpred

A data.frame with the predicted values from each sublearner algorithm for the rows in newx. If we have trained M individual models, then there will be M columns.

Details

If newx is omitted, the predicted values from object are returned. The learner algorithm needs to have a corresponding prediction function with ``predict'' prefixed onto the algorithm name (e.g. predict.SL.glm for SL.glm). This should be taken care of by the SuperLearner package.

See Also

subsemble

Examples

# NOT RUN {
# See subsemble() function documentation for an example.

# }