Energy statistics ($\mathcal{\varepsilon}$-statistics) are functions of distances between statistical observations. This class of functions has enabled the development of non-linear statistical concepts, termed distance variance, distance covariance, distance correlation, etc. The computational burden associated with the $\mathcal{\varepsilon}$-statistical quantities is really heavy and when the data reside in the multivariate space, the task becomes even harder. We alleviate this cost by tremendously reducing the memory requirements and essentially making the computations faster. We show the process for the cases of (univariate and multivariate) distance variance, distance covariance, (partial) distance correlation, energy distance and hypothesis testing for the equality of univariate distributions.
翻译:暂无翻译