brainconn.modularity.partition_distance

partition_distance(cx, cy)

This function quantifies the distance between pairs of community partitions with information theoretic measures.

Parameters:

• cx (Nx1 numpy.ndarray) – community affiliation vector X
• cy (Nx1 numpy.ndarray) – community affiliation vector Y

Returns:

• VIn (Nx1 numpy.ndarray) – normalized variation of information
• MIn (Nx1 numpy.ndarray) – normalized mutual information

Notes

(Definitions:

VIn = [H(X) + H(Y) - 2MI(X,Y)]/log(n) MIn = 2MI(X,Y)/[H(X)+H(Y)]

where H is entropy, MI is mutual information and n is number of nodes)