A modified version of the classification and regression tree (CART) algorithm for modelling spatial data that features coordinate information. Coordinate information can be used to evaluate measures of spatial autocorrelation and spatial compactness during the splitting phase of the tree, leading to better predictions and more physically realistic predictions on these types of datasets. These methods are described in Ancell and Bean (2021) <arXiv:2101.08258>.
| Version: | 1.4.5 |
| Depends: | fields, mgcv |
| Imports: | Rcpp, RcppParallel |
| LinkingTo: | Rcpp, RcppArmadillo, RcppParallel |
| Suggests: | testthat, knitr, rmarkdown |
| Published: | 2021-05-27 |
| Author: | Ethan Ancell [aut, cre] |
| Maintainer: | Ethan Ancell <ethanancell at gmail.com> |
| License: | MIT + file LICENSE |
| NeedsCompilation: | yes |
| SystemRequirements: | GNU make |
| Materials: | README NEWS |
| CRAN checks: | autocart results |
| Reference manual: | autocart.pdf |
| Vignettes: |
Introduction to the autocart package |
| Package source: | autocart_1.4.5.tar.gz |
| Windows binaries: | r-devel: autocart_1.4.5.zip, r-release: autocart_1.4.5.zip, r-oldrel: autocart_1.4.5.zip |
| macOS binaries: | r-release (arm64): autocart_1.4.5.tgz, r-oldrel (arm64): autocart_1.4.5.tgz, r-release (x86_64): autocart_1.4.5.tgz, r-oldrel (x86_64): autocart_1.4.5.tgz |
| Old sources: | autocart archive |
Please use the canonical form https://CRAN.R-project.org/package=autocart to link to this page.