Flow models & representations

These chapters apply the same map equation to richer ways of turning your data into flow — flow that carries memory, spans layers, tracks time, follows node attributes, or crosses two node types. The first three build on state nodes; if you have not met them yet, read State nodes and higher-order flow first.

• Memory and state networks: flow that depends on where the walker came from.

• Multilayer networks: a node living in several layers (relationship types) at once; the home of the inter-layer relax rate.

• Temporal networks: multilayer networks whose layers are time windows.

• Networks with metadata: folding node attributes into the objective.

• Bipartite networks: two node types, and how declaring the boundary changes the flow model.