Skip to main content

Digraph dynamics with heterogeneous vertices

Digraph panel data (directed edges between vertices observed at two or more time points) arise in the social sciences and other fields of scientific inquiry. A popular approach to model such data is to embed the discrete observations of the digraph into an unobserved, continuous-time Markov process. A large and appealing family of Markov probability models was considered by Snijders (2001). A substantial disadvantage associated with the mentioned family is the assumption that the vertices have identical weights in relevant digraph functions, which may be a doubtful assumption.