Interpretation Tools for Partial Least Squares Regression

Various kinds of plots (observations, variables, correlations, weights, regression coefficients and Variable Importance in the Projection) and aids to interpretation (coefficients, Q2, correlations, redundancies) for partial least squares regressions computed with the 'pls' package, following Tenenhaus (1998, ISBN:2-7108-0735-1).

morepls

Interpretation tools for PLS regression

morepls provides functions for the interpretation of PLS regressions.

The documentation is available here: https://nicolas-robette.frama.io/morepls/

Graphical functions :

• two-dimensional plot of observations
• two-dimensional plot of correlations between variables and components
• two-dimensional plot of variables (Y loadings and X projections)
• two-dimensional plot of supplementary variables
• two-dimensional plot of interaction between two supplementary variables
• two-dimensional plot of main and partial effects of a supplementary variable
• bar plot of regression coefficients
• bar plot of X variables weights
• bar plot of X variables VIPs

Statistical indicators :

• correlations between variables and components
• R2 and redundancies between variables and components
• Q2 and cumulative Q2 indexes
• raw and standardized coefficients

Installation

Execute the following code within R:

if (!require(devtools)){
    install.packages('devtools')
    library(devtools)
}
install_git("https://framagit.org/nicolas-robette/morepls")

Citation

To cite morepls in publications, use :

Robette N. (2025), morepls : Interpretation tools for PLS regression in R, version 0.2, https://nicolas-robette.frama.io/morepls/

References

Martens, H., Næs, T. (1989) Multivariate calibration. Chichester: Wiley.

Tenenhaus, M. (1998) La Regression PLS. Théorie et Pratique. Editions TECHNIP, Paris.

The image in the hex sticker is outrageously taken from https://moreplease.com/ from Iain Merrick.