Package: soilVAE
Type: Package
Title: Supervised Variational Autoencoder Regression via 'reticulate'
Version: 0.2.0
Authors@R: person("Hugo", "Rodrigues", email = "rodrigues.machado.hugo@gmail.com",
                  role = c("aut", "cre"))
Description: Supervised latent-variable regression for high-dimensional predictors
    such as soil reflectance spectra. The model uses an encoder-decoder neural
    network with a stochastic Gaussian latent representation regularized by a
    Kullback-Leibler term, and a supervised prediction head trained jointly with
    the reconstruction objective (Kingma and Welling (2014)
    <doi:10.48550/arXiv.1312.6114>). Predictors and response are standardised
    inside the model, so new data are scaled like the training data. The
    implementation interfaces R with a 'Python' deep-learning backend ('Keras'
    and 'TensorFlow') and provides utilities for installation, training with
    early stopping, tuning, prediction, latent embeddings, and saving models.
License: MIT + file LICENSE
Encoding: UTF-8
Imports: reticulate (>= 1.41), stats, utils
Suggests: knitr, rmarkdown, prospectr, pls, testthat (>= 3.0.0)
Config/testthat/edition: 3
VignetteBuilder: knitr
SystemRequirements: Python (>= 3.9); TensorFlow (>= 2.16); Keras (>= 3)
Config/roxygen2/version: 8.1.1
NeedsCompilation: no
Packaged: 2026-11-05 02:52:07 UTC; rodrigues.h
Author: Hugo Rodrigues [aut, cre]
Maintainer: Hugo Rodrigues <rodrigues.machado.hugo@gmail.com>
Repository: CRAN
Date/Publication: 2026-11-08 21:50:02 UTC
Built: R 4.5.2; ; 2026-11-09 21:29:37 UTC; unix
