grpselect: grpselect

In RAEN: Random Approximate Elastic Net (RAEN) Variable Selection Method

Description

This is the split step, where variable in subgroups are selected

Usage

grpselect(fgrp, x, y, B = 50, parallel = TRUE)

Arguments

fgrp	the variable group object from 'deCorr“
x	the predictor matrix
y	a dataframe of time to event and event status. The primary outcome status is coded 1, the secondary outcome as 2, etc. The censored is coded as 0.
B	the number of bootstraps
parallel	whether to use multiple cores for parallel computing. Default is TRUE.

Value

a list of

• fselect: Names of the selected variables.

• prob: the generalized ridge variable importance.

• weight: the inverse of the ridge variable importance.

RAEN documentation built on Feb. 22, 2021, 9:07 a.m.