Memory efficient data structures & algorithms on (geo) graphs

This software stands under Apache License 2

The following command will make a part of Germany routable:

cd core; ./run.sh unterfranken

1. it downloads 40MB, unzips it to 450MB and creates road-files for graphhopper (40MB)
2. it builds graphhopper
3. and runs some shortest path queries on it

When executing the command again, then the existing graphhopper road-files and jars will be used. So, the UI should pop up fast (~2 seconds). After the UI popped up you can drag to move the map or scroll to zoom like in ordinary maps apps. Click once to select a departure and another click to select the destination. Then a route should pop up like in this SHINY ;) image

Visualization of

• a bidirectional Dijkstra
• A*

If you want to import full Germany do:

cd core; ./run.sh germany

1. For Germany it takes approx 25 minutes for the import and roughly 1 minute for the ugly&slow UI to pop up.
2. At the moment the UI is a bit rough and simple so, don't click or drag too much as it takes some time for this large road network

• Simple Routing - full Germany
• Spatial Key
• Speed up your Quad-tree
• Spatial Hashtable
• Author@Twitter
• Lumeo - Implementing a Graph API via Lucene
• Cassovary - A production ready (?) in-memory & memory efficient Graph DB