Pearson's correlation coefficient in the theory of networks: A comment

Zafar Ahmed; Sachin Kumar

paper

Pearson's correlation coefficient in the theory of networks: A comment

arXiv:1803.06937

Abstract

In statistics, the Pearson correlation coefficient $r_{x, y}$ determines the degree of linear correlation between two variables and it is known that $- 1 \leq r_{x, y} \leq 1$ . In the theory of networks, a curious expression proposed in [PRL {\bf 89} 208701 (2002)] for degree-degree correlation coefficient $r_{j_{i}, k_{i}}, i \in [1, M]$ has been in use. We realize that the suggested form is the conventional Pearson's coefficient for ${(j_{i}, k_{i}), (k_{i}, j_{i})}$ for $2 M$ data points and hence it is rightly dedicated to undirected networks.

Important changes in the abstract and in the text