bp_kfold_VIP_analysis: K-fold bootstrap and permutation over PLS-VIP

In sipss/AlpsLCMS: Signal processing of LC-MS metabolomics

Description

Bootstrap and permutation over PLS-VIP on Alpslcms can be performed on both lcms_dataset_1D full spectra as well as lcms_dataset_peak_table peak tables.

Usage

bp_kfold_VIP_analysis(
  dataset,
  y_column,
  k = 4,
  ncomp = NULL,
  nbootstrap = 300,
  multi = NULL
)

Arguments

dataset	An lcms_dataset object
y_column	A list with the clases of the samples
k	Number of folds, recomended between 4 to 10
ncomp	number of components for the bootstrap models, if NULL it will be automatically calculated
nbootstrap	number of bootstrap dataset

Details

Use of the bootstrap and permutation methods for a more robust variable importance in the projection metric for partial least squares regression, in a k-fold cross validation

Value

A list with the following elements:

• important_vips: A list with the important vips selected

• relevant_vips: List of vips with some relevance

• wilcoxon_vips: List of vips that pass a wilcoxon test

• vip_means: Means of the vips scores

• vip_score_plot: plot of the vips scores

• kfold_resuls: results of the k bp_VIP_analysis

• kfold_index: list of index of partitions of the folds

sipss/AlpsLCMS documentation built on May 13, 2021, 6:18 p.m.