sp.logiregs: Many approximate simple logistic regressions.

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

Many approximate simple logistic regressions.

Description

Many approximate simple logistic regressions.

Usage

sp.logiregs(target, dataset, logged = FALSE)

Arguments

target	The dependent variable, a numerical vector with 0s or 1s.
dataset	A matrix with the indendent variables.
logged	Should the p-values be returned (FALSE) or their logarithm (TRUE)?

Details

Many simple approximate logistic regressions are performed and hypothesis testing for the singificance of each coefficient is returned. The code is available in the paper by Sikorska et al. (2013). We simply took the code and made some minor modifications. The explanation and the motivation can be found in their paper. They call it semi-parallel logistic regressions, hence we named the function sp.logiregs.

Value

A two-column matrix with the test statistics (Wald statistic) and their associated p-values (or their loggarithm).

Author(s)

Initial author Karolina Sikorska in the above reference paper. Modifications by Michail Tsagris.

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

References

Karolina Sikorska, Emmanuel Lesaffre, Patrick FJ Groenen and Paul HC Eilers (2013), 14:166. GWAS on your notebook: fast semi-parallel linear and logistic regression for genome-wide association studies. https://bmcbioinformatics.biomedcentral.com/track/pdf/10.1186/1471-2105-14-166.pdf

See Also

logiquant.regs, bic.regs

Examples

y <- rbinom(200, 1, 0.5)
x <- matrix( rnorm(200 * 30), ncol = 30 )
a <- MXM::sp.logiregs(y, x)

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