Implements Kolmogorov-Arnold Networks (KAN) whose edge functions are B-splines plus a residual base activation, written in native R on top of the 'torch' package. Provides Cox-de Boor basis evaluation with explicit extrapolation policies, multi-layer networks, Adam and L-BFGS training with validation, early stopping, input and output normalization, gradient clipping and reproducible seeds, adaptive (sample-based) grid updates, grid refinement, KAN-specific sparsity regularization, pruning, edge-function visualization and model persistence. A dependency-free base R reference implementation with analytic gradients is included as a numerical oracle and as a training backend that does not need 'libtorch'. The design follows Liu et al. (2024) <doi:10.48550/arXiv.2404.19756>.
| Version: | 0.5.1 |
| Depends: | R (≥ 4.1.0) |
| Imports: | graphics, grDevices, splines, stats, torch (≥ 0.13.0), utils |
| Suggests: | knitr, rmarkdown, testthat (≥ 3.0.0), jsonlite |
| Published: | 2026-11-17 |
| DOI: | 10.32614/CRAN.package.splineKANr (may not be active yet) |
| Author: | Thang Le-Dinh [aut, cre, cph] |
| Maintainer: | Thang Le-Dinh <thang87sg at gmail.com> |
| License: | MIT + file LICENSE |
| NeedsCompilation: | no |
| Language: | en-US |
| Materials: | README, NEWS |
| CRAN checks: | splineKANr results |
| Reference manual: | splineKANr.html , splineKANr.pdf |
| Vignettes: |
Method, validation and application (source, R code) Getting started with splineKANr (source, R code) |
| Package source: | splineKANr_0.5.1.tar.gz |
| Windows binaries: | r-devel: not available, r-release: not available, r-oldrel: not available |
| macOS binaries: | r-release (arm64): not available, r-oldrel (arm64): not available, r-release (x86_64): not available, r-oldrel (x86_64): not available |
Please use the canonical form https://CRAN.R-project.org/package=splineKANr to link to this page.