SuperLearner: summary.CV.SuperLearner – R documentation

Pricing

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

Get Started for Free

Documentation

SuperLearner

summary.CV.SuperLearner

Summary Function for Cross-Validated Super Learner

Description

summary method for the CV.SuperLearner function

Usage

## S3 method for class 'CV.SuperLearner'
summary(object, obsWeights = NULL, ...)

## S3 method for class 'summary.CV.SuperLearner'
print(x, digits, ...)

Arguments

`object`	An object of class "CV.SuperLearner", the result of a call to `CV.SuperLearner`.
`x`	An object of class "summary.CV.SuperLearner", the result of a call to `summary.CV.SuperLearner`.
`obsWeights`	Optional vector for observation weights.
`digits`	The number of significant digits to use when printing.
`...`	additional arguments ...

Details

Summary method for CV.SuperLearner. Calculates the V-fold cross-validated estimate of either the mean squared error or the -2*log(L) depending on the loss function used.

Value

summary.CV.SuperLearner returns a list with components

`call`	The function call from `CV.SuperLearner`
`method`	Describes the loss function used. Currently either least squares of negative log Likelihood.
`V`	Number of folds
`Risk.SL`	Risk estimate for the super learner
`Risk.dSL`	Risk estimate for the discrete super learner (the cross-validation selector)
`Risk.library`	A matrix with the risk estimates for each algorithm in the library
`Table`	A table with the mean risk estimate and standard deviation across the folds for the super learner and all algorithms in the library

Author(s)

Eric C Polley epolley@uchicago.edu