diffudist

Enables the evaluation of diffusion distances for complex single-layer networks. Given a network one can define different types of Laplacian (or transition) matrices corresponding to different continuous-time random walks dynamics on the network. This package enables the evaluation of Laplacians, stochastic matrices, and the corresponding diffusion distance matrices. The metric structure induced by the network-driven process is richer and more robust than the one given by shortest-paths and allows to study the geometry induced by different types of diffusion-like communication mechanisms taking place on complex networks. For more details see: De Domenico, M. (2017) doi:10.1103/physrevlett.118.168301 and Bertagnolli, G. and De Domenico, M. (2021) doi:10.1103/PhysRevE.103.042301.