variableSelection: SNPs selection

In MPAgenomics: Multi-Patient Analysis of Genomic Markers

Description

This function selects the most relevant variables according to a response.

Usage

variableSelection(
  dataMatrix,
  dataResponse,
  nbFolds = min(length(dataResponse), 10),
  loss = c("logistic", "linear"),
  plot = TRUE,
  pkg = c("HDPenReg", "spikeslab"),
  ...
)

Arguments

dataMatrix	Matrix containing the data, each row is a different sample.
dataResponse	response associated to the data.
nbFolds	number of folds in the cross validation.
loss	either "logistic" (binary response) or "linear" (quantitative response).
plot	If TRUE plot cross-validation mean squared error (default=TRUE).
pkg	Either "HDPenReg" or "spikeslab". Ued package in linear case.
...	spplementary arguments for cv.glmnet function in case of logistic loss or for HDlars or spikeslab function for linear loss.

Value

a list containing

variable

A vector containing the index of all selected variables.

coefficient

A vector containing the coefficients of all selected variables.

intercept

Intercept of the model.

Author(s)

Quentin Grimonprez

MPAgenomics documentation built on March 30, 2021, 5:13 p.m.