Package: cpfa
Type: Package
Title: Classification with Parallel Factor Analysis
Version: 1.2-8
Date: 2026-04-13
Authors@R: person(c("Matthew", "A."), "Asisgress", email = "mattgress@protonmail.ch", role = c("aut", "cre"))
Maintainer: Matthew A. Asisgress <mattgress@protonmail.ch>
Depends: multiway
Imports: glmnet, e1071, randomForest, nnet, rda, xgboost, foreach,
        doParallel, doRNG
Suggests: knitr, rmarkdown
Description: Classification using Richard A. Harshman's Parallel Factor Analysis-1 (Parafac) model or Parallel Factor Analysis-2 (Parafac2) model fit to a three-way or four-way data array. See Harshman and Lundy (1994): <doi:10.1016/0167-9473(94)90132-5>. Classification using principal component analysis (PCA) fit to a two-way data matrix is also supported. Uses component weights from one mode of a Parafac, Parafac2, or PCA model as features to tune parameters for one or more classification methods via a k-fold cross-validation procedure. Allows for constraints on different tensor modes. Allows for inclusion of additional features alongside features generated by the component model. Supports penalized logistic regression, support vector machine, random forest, feed-forward neural network, regularized discriminant analysis, and gradient boosting machine. Supports binary and multiclass classification. Predicts class labels or class probabilities, and calculates multiple classification performance measures. Implements parallel computing via the 'foreach', 'doParallel', and 'doRNG' packages.
License: GPL (>= 2)
VignetteBuilder: knitr, rmarkdown
NeedsCompilation: no
Packaged: 2026-04-12 18:51:19 UTC; nr2
Author: Matthew A. Asisgress [aut, cre]
Repository: CRAN
Date/Publication: 2026-04-14 00:50:03 UTC
Built: R 4.6.0; ; 2026-04-25 16:41:47 UTC; unix
