rbbnp: A Bias Bound Approach to Non-Parametric Inference

A novel bias-bound approach for non-parametric inference is introduced, focusing on both density and conditional expectation estimation. It constructs valid confidence intervals that account for the presence of a non-negligible bias and thus make it possible to perform inference with optimal mean squared error minimizing bandwidths. This package is based on Schennach (2020) <doi:10.1093/restud/rdz065>.

Version: 1.1.0
Depends: R (≥ 3.5)
Imports: dplyr, ggplot2 (≥ 3.4.0), gridExtra, pracma, purrr, Rglpk, tidyr
Suggests: knitr, rmarkdown, testthat (≥ 3.0.0)
Published: 2026-06-09
DOI: 10.32614/CRAN.package.rbbnp
Author: Xinyu DAI [aut, cre], Susanne M Schennach [aut]
Maintainer: Xinyu DAI <xinyu_dai at brown.edu>
License: GPL (≥ 3)
URL: https://xinyudai.net/rbbnp-dev/
NeedsCompilation: no
Citation: rbbnp citation info
Materials: README, NEWS
CRAN checks: rbbnp results

Documentation:

Reference manual: rbbnp.html , rbbnp.pdf
Vignettes: Density Estimation with rbbnp (source, R code)
Get Started with rbbnp (source, R code)
Regression with rbbnp (source, R code)
Theoretical Background (source, R code)

Downloads:

Package source: rbbnp_1.1.0.tar.gz
Windows binaries: r-devel: rbbnp_0.3.0.zip, r-release: rbbnp_0.3.0.zip, r-oldrel: rbbnp_0.3.0.zip
macOS binaries: r-release (arm64): rbbnp_1.1.0.tgz, r-oldrel (arm64): rbbnp_1.1.0.tgz, r-release (x86_64): rbbnp_1.1.0.tgz, r-oldrel (x86_64): rbbnp_1.1.0.tgz
Old sources: rbbnp archive

Linking:

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