leekasso: Leekasso regression

In 3inar/einr: Different Functions That I Keep Needing

Description

Jeff Leek's leekasso for p-value-based feature selection. Performs univariate regressions of the response on each of the predictors and selects the ten predictors with the lowest p-values. Then regressses the response on these ten predictors.

Usage

leekasso(x, y, family = gaussian)

Arguments

x	A predictor matrix with samples in the rows.
y	A response vector.
family	A description of the error distribution/link function for the glm. For more info see the documentation for glm.

Value

leekasso returns an object of class "leekasso" with the following components:

model	a glm object fit with the best ten predictors.
pvals	the p-values from the univariate regressions.
predictors	the indexes of the best predictors.
x	the data frame used to fit the model.
y	the responses used to fit the model.

See Also

glm

3inar/einr documentation built on May 5, 2019, 10:43 a.m.