Skip to contents

Extract optimal number of features to keep from cross-validation results for sO2PLS

Source: R/so2pls.R
so2pls_get_optim_keep.Rd

Extracts the optimal number of features to retain from datasets X and Y for the joint components.

Usage

so2pls_get_optim_keep(cv_res, use_1sd_rule = TRUE)

Arguments

cv_res

List, result from a call to the so2pls_crossval_sparsity() or OmicsPLS::crossval_sparsity().

use_1sd_rule

Boolean, should the 1 standard deviation rule be used when selecting the optimal number of features to retain? See Details.

Value

A list with elements keepx and keepy, each a vector of length equal to the number of joint components, where the ith element giving the number of features to retain from dataset X (keepx) or Y (keepy) for the i-th joint component.

Details

The 1-SD rule means that we are retaining the smallest number of features yielding an average covariance that is within 1SD of the maximum covariance obtained.