Greedy Spatial Navigation

Lee & Holme · PRL 108, 128701 (2012) — 100 real city road networks Demo by Claude Sonnet 4.6
computing…
City
— select a city
Interaction
Click a node → source.  Click another → target.
Hover to inspect intersections.
Algorithm
slow fast
Results
d_g — GSN steps— d — shortest path— d_r — random DFS—
ν = d / d_g— backtracks—
What is ν?
ν = d / d_g measures navigability: ratio of shortest-path length to greedy path length. ν→1 means the city layout allows near-optimal greedy routing. Removing some edges can increase ν (Braess's paradox).
PATH LOG
Select a city above, then click two nodes.