Tools for the Continuous Convolution Trick in Nonparametric Estimation

Implements the uniform scaled beta distribution and the continuous convolution kernel density estimator.

cctools

cctools is an R package implementing the uniform scaled beta distribution, a generic function for continuous convolution, and the continuous convolution kernel density estimator, see Nagler (2017).

How to install

• the stable release on CRAN:

install.packages("cctools")

• the latest development version:

devtools::install_github("tnagler/cctools")

Functions

For a detailed description, see the API documentation.

• dusb(), rusb(): Density and simulation functions for the uniform scaled beta distribution.
• cont_conv(): Expands all factor variables in a data frame and applies the continuous convolution tricks to all ordered() variables.
• cckde(), dcckde(), predict.cckde(): fit and evaluate the continuous convolution kernel density estimator.

References

Nagler, T. (2017). A generic approach to nonparametric function estimation with mixed data. arXiv:1704.07457