cor.drop1: Drop all possible single terms from a model using the partial...

In MXM: Feature Selection (Including Multiple Solutions) and Bayesian Networks

Drop all possible single terms from a model using the partial correlation

Description

Drop all possible single terms from a model using the partial correlation.

Usage

cor.drop1(y, x, logged = FALSE)

Arguments

y	A numerical vector with the response variable.
x	A numerical matrix or a data.frame with the predictor variables. If is is a matrix it is internally transformed into a data.frame form, hence the user is advised to supply a data.frame in order to save some time. If the number of columns (variables) is higher than the number of rows (observations) the function will simply not work.
logged	If you want the p-values be returned leave this FALSE. If it is TRUE their logarithm is returned.

Details

This uses R's command drop1 and modifies it so as to calculate the p-value using Fisher's conditional independence test.

Value

A matrix with two columns, the test statistic values and its associated p-value.

Author(s)

Michail Tsagris

R implementation and documentation: Michail Tsagris [e-mail]

See Also

glm.bsreg, fbed.reg, mmpcbackphase

Examples

y <- rnorm(200)
x <- matrix( rnorm(200 * 10), ncol = 10)
cor.drop1(y, x)

MXM documentation built on Aug. 25, 2022, 9:05 a.m.