City2Graph

The City2Graph tab builds heterogeneous urban graphs from real-world data: buildings, streets, amenities, and transit, connected by morphological, proximity, and metapath relationships. It targets urban analytics and GeoAI workflows.

The root City2Graph panel is a status header. It reports the active object’s type (OSMnx network, City2Graph features, tessellation, morphological graph, or travel graph) and its feature/node/edge counts. OSMnx street networks can be used directly by every City2Graph operation, so the two tabs are designed to be used together.

Panels in this tab

Each panel of the addon has its own reference page:

Panel What it does
Data Import Acquire the raw layers: Overture/OSMnx features, local files, GTFS transit.
Urban Morphology Build tessellations and morphological building/street graphs.
Transportation Turn a loaded GTFS feed into travel-summary and OD graphs.
Mobility (OD Matrix) Build graphs from an external origin–destination matrix.
Graph Tools Clip, filter, isochrone, and largest-component utilities.
Proximity Graphs Spatial networks (KNN, radius, Delaunay…) over feature objects.
Metapath Analysis Connect amenities through the street network.
GeoAI Export Export graphs for GNN training (JSON/GraphML, PyG loader).

Header status fields

When a compatible object is active, the header reports:

Field Meaning
Object Name of the active object.
Features Feature count for City2Graph feature objects.
Nodes / Edges Node and edge counts for graph objects.
Type Detected type: OSMnx Street Network, Urban Tessellation, Morphological Graph, or Travel Graph.

If nothing compatible is selected, the header invites you to import data with the panels below.

Typical workflow

  1. In Data Import, define an area and download buildings, segments, and places; load a GTFS feed if you need transit.
  2. Build a Morphological Graph, generate Proximity Graphs over amenities, or run Metapath Analysis to connect amenities along the street network. For transit, use Transportation or Mobility.
  3. Refine the result with Graph Tools (keep the largest component, clip, isochrone).
  4. Visualize in Blender (city2graph layers carry per-type attributes) and/or export for GNN training.
Tip

Most operations align their output with the active OSMnx graph when one is present, so downloaded features, dual graphs, bridges, and metapaths all share the same coordinate space as the street network.

Back to top