Python Data Analysis Cookbook

Estimating the average clustering coefficient

From kindergarten onward, we have friends, close friends, best friends forever, social media friends, and other friends. A social network graph should, therefore, have clumps, unlike what you would observe at a high school party. The question that naturally arises is what would happen if we just invite a group of random strangers to a party or recreate this setup online? We would expect the probability of strangers connecting to be lower than for friends. In graph theory, this probability is measured by the clustering coefficient.

The average clustering coefficient is a local (single node) version of the clustering coefficient. The definition of this metric considers triangles formed by nodes. With ...