Package: lwpls
Type: Package
Title: Locally-Weighted Partial Least Squares for 'tidymodels'
Version: 0.3.0
Authors@R: c(
    person("Christian L.", "Goueguel", , "christian.goueguel@gmail.com", role = c("aut", "cre")),
    person("Hiromasa", "Kaneko", role = "cph",
           comment = "reference LW-PLS implementation transcribed in the tests")
  )
Description: Locally-weighted partial least squares (LW-PLS) is a
    just-in-time learning method that builds a dedicated weighted partial
    least squares model around every new sample, where training samples
    are weighted by their similarity to that sample (Kim et al. (2011)
    <doi:10.1016/j.ijpharm.2011.10.007>). Regression (single or multiple
    outcomes) and classification (locally weighted partial least
    squares-discriminant analysis; Bevilacqua and Marini (2014)
    <doi:10.1016/j.aca.2014.05.057>) are supported. The local models can
    be robust to outliers, by partial robust M-regression (Serneels et al.
    (2005) <doi:10.1016/j.chemolab.2005.04.007>), and sparse (Hoffmann et
    al. (2015) <doi:10.1016/j.chemolab.2015.09.019>). The prediction kernel
    is written in C++ with 'RcppArmadillo', and the model is integrated
    with the 'parsnip', 'dials', 'tune' and 'yardstick' packages so it can
    be used within the 'tidymodels' framework. Diagnostics of the local
    models (prediction reliability, local regression vectors, robust
    weights and selected predictors) come with 'ggplot2' plots designed for
    spectroscopic data.
License: MIT + file LICENSE
URL: https://github.com/ChristianGoueguel/lwpls,
        https://christiangoueguel.com/lwpls/
BugReports: https://github.com/ChristianGoueguel/lwpls/issues
Depends: R (>= 4.1.0)
Imports: cli, dials (>= 1.2.0), generics, ggplot2 (>= 3.5.0), hardhat
        (>= 1.3.0), parsnip (>= 1.3.0), Rcpp, rlang (>= 1.1.0), scales
        (>= 1.3.0), stats, tibble, vctrs, yardstick (>= 1.2.0)
Suggests: dplyr, knitr, modeldata, plotly, pls, recipes, rmarkdown,
        rsample, spelling, testthat (>= 3.0.0), tune (>= 1.1.0), withr,
        workflows
LinkingTo: Rcpp, RcppArmadillo
VignetteBuilder: knitr
Config/testthat/edition: 3
Encoding: UTF-8
Language: en-US
Config/roxygen2/version: 8.1.0
NeedsCompilation: yes
Packaged: 2026-10-07 19:35:25 UTC; christiangoueguel
Author: Christian L. Goueguel [aut, cre],
  Hiromasa Kaneko [cph] (reference LW-PLS implementation transcribed in
    the tests)
Maintainer: Christian L. Goueguel <christian.goueguel@gmail.com>
Repository: CRAN
Date/Publication: 2026-10-18 15:00:02 UTC
