Skip to content

Latest commit

 

History

3 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 

Repository files navigation

GraphFlow

Where traffic flows, and where it stops.
Road networks · Routing · ML-based congestion prediction

status python license


What is GraphFlow?

GraphFlow is not just another shortest-path library. It answers a harder question:

"What will the road look like in an hour?"

Classical routing tells you the shortest path now. GraphFlow predicts how traffic will change, which edges will jam, and how to reroute before you hit congestion.

Built on the intersection of graph algorithms and machine learning.


Features

  • Road network loading from OpenStreetMap (RoadGraph.from_osm())
  • Classical routing — Dijkstra, A*, bidirectional search
  • Edge speed prediction — ML model trained on road features
  • Congestion forecasting — predict which edges will slow down
  • Dynamic rerouting — recalculate path when traffic changes
  • Arrival time uncertainty — not just ETA, but confidence interval
  • Visualization — network, routes, congestion heatmaps

Tech Stack

  • Python 3.10+
  • NetworkX / igraph — graph structures
  • OSMnx — OpenStreetMap integration
  • NumPy, scikit-learn — ML models
  • Matplotlib / Plotly — visualization

Roadmap

  • Core graph and routing algorithms
  • Edge speed prediction (ML)
  • Congestion-aware dynamic rerouting
  • Arrival time uncertainty estimation
  • Route and flow visualization
  • Publication and documentation

Science

GraphFlow sits at the intersection of:

  • Graph theory — flow networks, min-cut, max-flow
  • Spatial ML — edge features, node embeddings
  • Time series — traffic patterns, periodicity
  • Uncertainty quantification — ETA confidence intervals

This is not just Dijkstra. This is routing under uncertainty.


Author

Emelyanov Ilya GitHub: @nsdmlk


License

MIT © Emelyanov Ilya, 2026


Roads are graphs. Traffic is data. GraphFlow connects them.

About

Road network analysis that goes beyond shortest paths. Combines graph algorithms with ML to predict congestion, reroute dynamically, and estimate arrival time uncertainty. Built for routing under uncertainty.

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages