dcor: Distance correlation and energy statistics in Python
EntityUAM. Departamento de Matemáticas
10.1016/j.softx.2023.101326SoftwareX 22 (2023): 101326
Funded byPID2019-106827GB-I00, PID2019-106827GB-I00
ProjectGobierno de España. PID2019–106827GB-I00; Gobierno de España. PID2019-109387GB-I00
SubjectsHypothesis testing; High Level languages; Vectorization; Distance Correlation; Matemáticas
Rights© 2023 The Authors
Esta obra está bajo una Licencia Creative Commons Atribución 4.0 Internacional.
This article presents dcor, an open-source Python package dedicated to distance correlation and other statistics related to energy distance. These energy statistics include distances between distributions and the associated tests for homogeneity and independence. Some of the most efficient algorithms for the estimation of these measures have been implemented relying on optimization techniques such as vectorization, compilation, and parallelization. The performance of these estimators is evaluated by comparison with alternative implementations in other packages. The package is also designed to be compatible with the packages conforming the scientific Python ecosystem. With that purpose in mind, dcor is an early adopter of the Python array API standard.
This item appears in the following Collection(s)
Showing items related by title, author, creator and subject.