English

Pearson Distance is not a Distance

Methodology 2019-08-19 v1 Machine Learning

Abstract

The Pearson distance between a pair of random variables X,YX,Y with correlation ρxy\rho_{xy}, namely, 1-ρxy\rho_{xy}, has gained widespread use, particularly for clustering, in areas such as gene expression analysis, brain imaging and cyber security. In all these applications it is implicitly assumed/required that the distance measures be metrics, thus satisfying the triangle inequality. We show however, that Pearson distance is not a metric. We go on to show that this can be repaired by recalling the result, (well known in other literature) that 1ρxy\sqrt{1-\rho_{xy}} is a metric. We similarly show that a related measure of interest, 1ρxy1-|\rho_{xy}|, which is invariant to the sign of ρxy\rho_{xy}, is not a metric but that 1ρxy2\sqrt{1-\rho_{xy}^2} is. We also give generalizations of these results.

Keywords

Cite

@article{arxiv.1908.06029,
  title  = {Pearson Distance is not a Distance},
  author = {Victor Solo},
  journal= {arXiv preprint arXiv:1908.06029},
  year   = {2019}
}
R2 v1 2026-06-23T10:49:14.780Z