Fast Computation of Distance Correlations

Fast computation of the distance covariance 'dcov' and distance correlation 'dcor'. The computation cost is only O(n log(n)) for the distance correlation (see Chaudhuri, Hu (2019) ). The functions are written entirely in C++ to speed up the computation.

The dccpp R Package

Fast computation of the distance covariance dcov and distance correlation dcor. The computation cost is only O(n log(n)) for the distance correlation (see Chaudhuri, Hu, 2019, arXiv, elsevier). The functions are written entirely in C++ to speed up the computation.

Installation

Install from CRAN

install.packages("dccpp")

Install from GitHub

# install.packages("remotes")
remotes::install_github("BerriJ/dccpp")

Usage

dcov(x, y) # To calculate the distance covariance
dcor(x, y) # To calculate the distance correlation

Contributions and Issues

Feel free to raise an issue if you find something not working properly.

You are also very welcome to contribute to dcccp. Please base your pull requests on the development branch. Note that this package focuses on performance, PR's that improve the performance are particularly welcome.