Maps of real-world cities in graph representation (with nodes and edges)

Generate graph representations (with nodes and edges) of real-world cities:

The goal of the project is to provide maps of real cities in a graph representation (with nodes and edges). Nodes are intersections between roads, points where roads change direction, or subway stations. We create an edge between two nodes whenever there is a road or railway that links them. If we zoom in Washington Square, NYC, this is what it looks like:

Check out the examples of NYC, Boston, and the amazing Madrid:

The resulting graphs are big! For instance, the graph representation of NYC counts 98542 nodes and 114981 edges. For simplicity, we consider railways and roads separately.

City	Mapzen - OpenStreetMap	Snapshot of the roads	Donwload PDF	Donwload gexf file
New York City			Roads PDF Railways PDF	Roads GEXF Railways GEXF
Boston			Roads PDF Railways PDF	Roads GEXF Railways GEXF
Madrid			Roads PDF Railways PDF	Roads GEXF Railways GEXF

Tutorial

Step 1: Download geoJSON data

Download the geoJSON data of one of the cities available at mapzen.com

Step 2: Download the graphFromGeoJson.jar file from here

Step 3: Generate the graph

Six arguments need to be specified:

1. path_to_map: path to the geojson file

2. road_class: highway or railway

3. minX: minimum x coordinate of the rectangle delimiting the city

4. maxX: maximum x coordinate of the rectangle delimiting the city

5. minY: minimum y coordinate of the rectangle delimiting the city

6. maxY: maximum y coordinate of the rectangle delimiting the city

From the terminal:

$ java -jar graphFromGeoJson.jar path_to_map road_class minX maxX minY maxY > output.gexf

Example

$ java -jar graphFromGeoJson.jar new-york_new-york-roads.geojson railway -74.0969467163086 -73.8339614868164 40.70016219564594 40.85147526676901 > nyc_railway.gefx

Step 4: Visualization

We use Gephi for the visualizations.

Authors and Contributors

This project was developed by Ignacio Arnaldo (@ignacioarnaldo)