PLSR: Partial least squares regression

In computational-metabolomics/structtoolbox: Data processing & analysis tools for Metabolomics and other omics

Partial least squares regression

Description

PLS is a multivariate regression technique that extracts latent variables maximising covariance between the input data and the response. For regression the response is a continuous variable.

Usage

PLSR(number_components = 2, factor_name, ...)

Arguments

number_components	(numeric, integer) The number of PLS components. The default is 2.
factor_name	(character) The name of sample meta column(s) to use.
...	Additional slots and values passed to struct_class.

Details

This object makes use of functionality from the following packages:

• pls

Value

A PLSR object with the following output slots:

scores	(DatasetExperiment)
loadings	(data.frame)
yhat	(data.frame)
y	(data.frame)
reg_coeff	(data.frame)
vip	(data.frame)
pls_model	(list)
pred	(data.frame)
sr	(data.frame) Selectivity ratio for a variable represents a measure of a variable's importance in the PLS model. The output data.frame contains a column of selectivity ratios, a column of p-values based on an F-distribution and a column indicating significance at p < 0.05.
sr_pvalue	(data.frame) A p-value computed from the Selectivity Ratio based on an F-distribution.

Inheritance

A PLSR object inherits the following struct classes:

⁠[PLSR]⁠ >> ⁠[model]⁠ >> ⁠[struct_class]⁠

References

Liland K, Mevik B, Wehrens R (2023). pls: Partial Least Squares and Principal Component Regression. R package version 2.8-3, https://CRAN.R-project.org/package=pls.

Examples

M = PLSR(
      number_components = 2,
      factor_name = "V1")

M = PLSR(factor_name='run_order')

computational-metabolomics/structtoolbox documentation built on Feb. 9, 2024, 8:19 a.m.