gkrreg: Gaussian Kernel Robust Regression (GKRReg)

Implements the Gaussian Kernel Robust Regression (GKRReg / GKRR) method proposed by De Carvalho, Lima Neto and Ferreira (2017) <doi:10.1016/j.neucom.2016.12.035>. The method re-weights observations iteratively using the Gaussian kernel so that poorly-fitted observations (outliers, leverage points) receive small weights, yielding resistance to Y-space outliers, X-space outliers and leverage points. Convergence is guaranteed by Propositions 4.1 and 4.2 of the original paper. Three estimators for the kernel width hyper-parameter are provided (S1: Caputo, S2: pairwise median, S3: residual variance). Inference is provided via an analytic sandwich variance estimator (default) or via bootstrap (percentile, normal and BCa intervals with p-values) through gkrr_boot(). Six real datasets from the robust regression literature are included to facilitate reproducible comparisons.

Version: 0.4.0
Depends: R (≥ 4.0.0)
Imports: stats, graphics, grDevices, MASS, sm
Suggests: robustbase, quantreg, testthat (≥ 3.0.0), knitr, rmarkdown
Published: 2026-06-17
DOI: 10.32614/CRAN.package.gkrreg (may not be active yet)
Author: Eufrásio de Andrade Lima Neto [aut], Marcelo Rodrigo Portela Ferreira [aut, cre]
Maintainer: Marcelo Rodrigo Portela Ferreira <marcelo at de.ufpb.br>
BugReports: https://github.com/marcelorpf/gkrreg/issues
License: GPL-3
URL: https://github.com/marcelorpf/gkrreg
NeedsCompilation: no
Materials: NEWS
CRAN checks: gkrreg results

Documentation:

Reference manual: gkrreg.html , gkrreg.pdf
Vignettes: Introduction to gkrreg: Gaussian Kernel Robust Regression (source, R code)

Downloads:

Package source: gkrreg_0.4.0.tar.gz
Windows binaries: r-devel: not available, r-release: not available, r-oldrel: not available
macOS binaries: r-release (arm64): gkrreg_0.4.0.tgz, r-oldrel (arm64): gkrreg_0.4.0.tgz, r-release (x86_64): gkrreg_0.4.0.tgz, r-oldrel (x86_64): gkrreg_0.4.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=gkrreg to link to this page.