Divisibility D is defined as:(3)where Ci indicates the community to which the node I belongs (here we can have either Ci = B1 or Ci = B2); the δ function yields 1 if vertices i and j are in the same community (i.e. in the same brain), and 0 otherwise; k is a positive constant (here set equal to W) to avoid possible divergence of D. The divisibility D is actually the inverse of the cut size [37] extended to weighted graphs.
