Concepts

These chapters teach the ideas behind Infomap from the ground up, in reading order. Start here if you want to understand why Infomap finds the communities it does before you run it on your own data.

• Flow and random walks: what “flow” means and why Infomap models a random walker.

• The map equation: the objective Infomap minimises (the codelength, the bits needed to describe that walk) and the partition that makes it shortest.

• Hierarchy and the multilevel map equation: how Infomap discovers nested structure to any depth, with no resolution parameter to tune.

• State nodes and higher-order flow: the mechanism behind the memory, multilayer, and temporal models; read it before the Flow models section.

• Reading Infomap through Louvain and Leiden: understand Infomap in the terms you already use for modularity.