pc.sel: Variable selection using the PC-simple algorithm

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

Variable selection using the PC-simple algorithm

Description

Variable selection using the PC-simple algorithm.

Usage

pc.sel(target, dataset, threshold = 0.05)

Arguments

target	A numerical vector with continuous data.
dataset	A matrix with numerical data; the independent variables, of which some will probably be selected.
threshold	The significance level.

Details

Variable selection for continuous data only is performed using the PC-simple algorithm (Buhlmann, Kalisch and Maathuis, 2010). The PC algorithm used to infer the skeleton of a Bayesian Network has been adopted in the context of variable selection. In other words, the PC algorithm is used for a single node.

Value

A list including:

vars	A vector with the selected variables.
n.tests	The number of tests performed.
runtime	The runtime of the algorithm.

Author(s)

Michail Tsagris

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

References

Buhlmann P., Kalisch M. and Maathuis M. H. (2010). Variable selection in high-dimensional linear models: partially faithful distributions and the PC-simple algorithm. Biometrika, 97(2), 261-278. https://arxiv.org/pdf/0906.3204.pdf

See Also

pc.skel, omp

Examples

y <- rnorm(100)
x <- matrix( rnorm(100 * 30), ncol = 30)
a <- MXM::pc.sel(y, x)
b <- MMPC(y, x)

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