predict.cvma: Get predictions on cvma object

In benkeser/cvma: Cross-validation-based maximal association

Description

Get predictions on cvma object

Usage

## S3 method for class 'cvma'
predict(object, newdata, outer = TRUE, ...)

Arguments

object	Object of class cvma
newdata	A data.frame of predictors on which to obtain predictions
outer	Return a vector of predictions from outer super learner (i.e., the one fit on all V folds, default) or a matrix of predictions from inner super learners fit on V-1 folds.
...	Other options (not currently used)

Value

A list of predictions with named entries y_weight and object$y_names. The former contains predictions of the weighted outcome, while the latter contains predictions of the respective univariate outcome. Each entry in the outputted list is itself a list with two entries. The first is a vector of super learner predictions for the particular outcome, while the second is a matrix with columns corresponding the the various learners predictions of the particular outcome. TO DO: Rethink formatting of output when outer = TRUE to be more like the format when outer = FALSE?

Examples

# TO DO: Add examples here

benkeser/cvma documentation built on May 5, 2019, 1:37 p.m.