Boundary Adaptive Local Polynomial Conditional Density Estimator

Tools for estimation and inference of conditional densities, derivatives and functions. This is the companion software for Cattaneo, Chandak, Jansson and Ma (2024) .

Local Polynomial Conditional Density Methods

The lpcde package implements bandwidth selection, point estimation, and inference procedures for local polynomial conditional distribution and density methods.

lpcde: local polynomial conditional CDF, PDF, and derivative estimation with pointwise and uniform inference.
lpbwcde: rule-of-thumb bandwidth selection for local polynomial conditional density estimation.

R Implementation

To install/update in R type:

install.packages('lpcde')

Help: CRAN repository.
Replication: R-script, software article replication, comparison illustration, Python illustration, Stata illustration.
Development version:

devtools::install_github('nppackages/lpcde/R/lpcde')

Basic usage:

model1 <- lpcde::lpcde(x_data = x_data, y_data = y_data,
                       y_grid = y_grid, x = 0, bw = 1)

model2 <- lpcde::lpbwcde(y_data = y_data, x_data = x_data,
                         x = 0, y_grid = y_grid)

Standard R methods including coef, confint, plot, print, summary, and vcov are available for fitted lpcde objects. The summary, print, and coef methods are available for lpbwcde bandwidth-selection objects.

References

Software and Implementation

Cattaneo, Chandak, Jansson and Ma (2025): lpcde: Estimation and Inference for Local Polynomial Conditional Density Estimators.
Journal of Open Source Software 10(107): 7241.
Companion arXiv version.

Technical and Methodological

Cattaneo, Chandak, Jansson and Ma (2024): Boundary Adaptive Local Polynomial Conditional Density Estimators.
Bernoulli 30(4): 3193-3223.
Supplemental appendix.

Funding

This work was supported in part by the National Science Foundation through grants SES-1947805, SES-1947662, DMS-2210561, and SES-2241575, and by the National Institutes of Health through grant R01 GM072611-16.