Ensemble Partial Least Squares Regression

An algorithmic framework for measuring feature importance, outlier detection, model applicability domain evaluation, and ensemble predictive modeling with (sparse) partial least squares regressions.

enpls offers an algorithmic framework for measuring feature importance, outlier detection, model applicability domain evaluation, and ensemble predictive modeling with (sparse) partial least squares regressions.

Installation

Install enpls from CRAN:

install.packages("enpls")

Or try the development version on GitHub:

# install.packages("devtools")
devtools::install_github("road2stat/enpls")

To load the package in R, simply use

library("enpls")

and you are all set. See the vignette (or open with vignette("enpls") in R) for a quick-start guide.

News

CHANGES IN enpls VERSION 5.6 (2016-11-25)

NEW FEATURES

New argument cvfolds now available in all applicable functions for finer control of cross-validation folds in automatic parameter selection of each PLS/SPLS model.

IMPROVEMENTS

Critical implementation improvements for processing the case where argument maxcomp = NULL (maximum number of components not specified explicitly) in enpls. functions. Now it should correctly determine the maximum number of components to use, considering both cross-validation and special cases such as n < p. Thanks to Dr. You-Wu Lin for the feedback.

CHANGES IN enpls VERSION 5.0 (2016-10-20)

NEW FEATURES

New functions enpls.ad() and enspls.ad() for PLS and sparse PLS model applicability domain evaluation.
New plot functions plot.enpls.ad() and plot.enpls.ad() for exploring model applicability domain evaluation results with traditional static plot support and interactive plot support.
New argument alpha available for setting transparency level (to reduce overplotting) in plot.cv.enpls, plot.enpls.od, plot.cv.enspls, and plot.enspls.od.

CHANGES IN enpls VERSION 4.5 (2016-09-15)

IMPROVEMENTS

Reduced memory footprints for enpls.fit() and enspls.fit().

NEW FEATURES

New functions enpls.rmse(), enpls.mae(), and enpls.rmsle() for computing RMSE, MAE, and RMSLE.

CHANGES IN enpls VERSION 4.0 (2016-08-28)

IMPROVEMENTS

General improvements on function documentation.
Changing the argument name MCtimes to reptimes.
Changing the option name "bootstrap" to "boot".

CHANGES IN enpls VERSION 3.0 (2016-06-22)

NEW FEATURES

Add sparse partial least squares regression.

IMPROVEMENTS

Improvements on documentation; rewritten vignette.

CHANGES IN enpls VERSION 2.0 (2016-06-19)

NEW FEATURES

New project website: http://enpls.org

IMPROVEMENTS

General improvements on plotting functions.

BUG FIXES

Fixed major bugs in cv.enpls and plotting functions.
Many other bug fixes.

CHANGES IN enpls VERSION 1.1 (2015-11-26)

BUG FIXES

Fixed the bugs in automatic component number selection which could make the intercept-only model rank best. Thanks for the test and patch from Max Kuhn <max.kuhn@pfizer.com>.