dcor

dcor: distance correlation and energy statistics in Python.

E-statistics are functions of distances between statistical observations in metric spaces.

Distance covariance and distance correlation are dependency measures between random vectors introduced in [SRB07] with a simple E-statistic estimator.

This package offers functions for calculating several E-statistics such as:

• Estimator of the energy distance [SR13].
• Biased and unbiased estimators of distance covariance and distance correlation [SRB07].
• Estimators of the partial distance covariance and partial distance covariance [SR14].

It also provides tests based on these E-statistics:

• Test of homogeneity based on the energy distance.
• Test of independence based on distance covariance.

Installation

dcor is on PyPi and can be installed using pip:

pip install dcor

It is also available for conda using the conda-forge channel:

conda install -c conda-forge dcor

Previous versions of the package were in the vnmabus channel. This channel will not be updated with new releases, and users are recommended to use the conda-forge channel.

Requirements

dcor is available in Python 3.8 or above in all operating systems. The package dcor depends on the following libraries:

• numpy
• numba >= 0.51
• scipy
• joblib

Citing dcor

Please, if you find this software useful in your work, reference it citing the following paper:

@article{ramos-carreno+torrecilla_2023_dcor,
  author = {Ramos-Carreño, Carlos and Torrecilla, José L.},
  doi = {10.1016/j.softx.2023.101326},
  journal = {SoftwareX},
  month = {2},
  title = {{dcor: Distance correlation and energy statistics in Python}},
  url = {https://www.sciencedirect.com/science/article/pii/S2352711023000225},
  volume = {22},
  year = {2023},
}

You can additionally cite the software repository itself using:

@misc{ramos-carreno_2022_dcor,
  author = {Ramos-Carreño, Carlos},
  doi = {10.5281/zenodo.3468124},
  month = {3},
  title = {dcor: distance correlation and energy statistics in Python},
  url = {https://github.com/vnmabus/dcor},
  year = {2022}
}

If you want to reference a particular version for reproducibility, check the version-specific DOIs available in Zenodo.

Documentation

The documentation can be found in https://dcor.readthedocs.io/en/latest/?badge=latest

References

[SR13]

Gábor J. Székely and Maria L. Rizzo. Energy statistics: a class of statistics based on distances. Journal of Statistical Planning and Inference, 143(8):1249 – 1272, 2013. URL: http://www.sciencedirect.com/science/article/pii/S0378375813000633, doi:10.1016/j.jspi.2013.03.018.

[SR14]

Gábor J. Székely and Maria L. Rizzo. Partial distance correlation with methods for dissimilarities. The Annals of Statistics, 42(6):2382–2412, 12 2014. doi:10.1214/14-AOS1255.

[SRB07]

(1, 2) Gábor J. Székely, Maria L. Rizzo, and Nail K. Bakirov. Measuring and testing dependence by correlation of distances. The Annals of Statistics, 35(6):2769–2794, 12 2007. doi:10.1214/009053607000000505.