{"name":"caretSBF","title":"Selection By Filtering (SBF) Helper Functions","pagetitle":"Selection By Filtering (SBF) Helper Functions — caretSBF","aliases":["caretSBF","lmSBF","rfSBF","treebagSBF","ldaSBF","nbSBF","gamScores","anovaScores","anovaScores","gamScores"],"author":"\nMax Kuhn\n","keywords":"models","description":"

Ancillary functions for univariate feature selection

","usage":"caretSBF\n\nanovaScores(x, y)\n\ngamScores(x, y)","arguments":[{"name":"x","description":"

a matrix or data frame of numeric predictors

"},{"name":"y","description":"

a numeric or factor vector of outcomes

"}],"has_args":true,"sections":[{"title":"Format","contents":"

An object of class list of length 5.

"}],"details":["

More details on these functions can be found at\nhttp://topepo.github.io/caret/feature-selection-using-univariate-filters.html.

This page documents the functions that are used in selection by filtering\n(SBF). The functions described here are passed to the algorithm via the\nfunctions argument of sbfControl.

See sbfControl for details on how these functions should be\ndefined.","anovaScores and gamScores are two examples of univariate\nfiltering functions. anovaScores fits a simple linear model between a\nsingle feature and the outcome, then the p-value for the whole model F-test\nis returned. gamScores fits a generalized additive model between a\nsingle predictor and the outcome using a smoothing spline basis function. A\np-value is generated using the whole model test from\nsummary.Gam and is returned.

If a particular model fails for lm or gam, a p-value of 1 is\nreturned.

"],"seealso":"

sbfControl, sbf,\nsummary.Gam

","package":{"package":"caret","version":"6.0-80"}}

Description

Usage

Arguments

Format

Details

See Also