PySGN: A Python package for constructing synthetic geospatial networks

Synthetic networks are commonly used to study the structure and dynamics of social systems, transportation infrastructure and other complex phenomena.Classical random graph models, such as the Erdős-Rényi, Watts-Strogatz and Barabási-Albert models, generate abstract networks with different structural characteristics: the Erdős-Rényi model connects nodes at random with equal probability, the Watts-Strogatz model rewires a ring lattice to produce small-world networks, and the Barabási-Albert model yields scale-free networks through preferential attachment (Barabási & Albert, 1999;Erdös & Rényi, 1960;Watts & Strogatz, 1998).In their standard form these models ignore the spatial positions of nodes; yet in many empirical settings (e.g., human social networks, commuting patterns or infrastructure networks) proximity strongly influences who connects to whom.