lspartition: Nonparametric Estimation and Inference Procedures using Partitioning-Based Least Squares Regression

Tools for statistical analysis using partitioning-based least squares regression as described in Cattaneo, Farrell and Feng (2018) <arXiv:1804.04916>. lsprobust() for nonparametric point estimation of regression functions and derivatives thereof, and for robust bias-corrected (pointwise and uniform) inference procedures. lspkselect() for data-driven procedure for selecting the IMSE-optimal number of knots. lsprobust.plot() for regression plots with robust confidence intervals and confidence bands. lsplincom() for estimation and inference for linear combinations of regression functions from different groups.

Version: 0.1
Depends: R (≥ 3.1)
Imports: ggplot2, pracma, mgcv, combinat, matrixStats, MASS
Published: 2018-04-19
Author: Matias D. Cattaneo, Max H. Farrell, Yingjie Feng
Maintainer: Yingjie Feng <yjfeng at>
License: GPL-2
NeedsCompilation: no
CRAN checks: lspartition results


Reference manual: lspartition.pdf
Package source: lspartition_0.1.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X binaries: r-release: lspartition_0.1.tgz, r-oldrel: lspartition_0.1.tgz


Please use the canonical form to link to this page.