Skip to main content

GraphML Format

Description​

GraphML is an XML format for attributed graphs used in network analysis and visualization tools. IterableData reads GraphML via NetworkX and yields node records first, then edge records. It is read-only in this release.

File Extensions​

  • .graphml — GraphML documents

Implementation Details​

Reading​

  • Parses with networkx.read_graphml
  • Converts the graph to a flat list: nodes then edges
  • Node rows: {"_type": "node", "id": "...", ...attributes}
  • Edge rows: {"_type": "edge", "source": "...", "target": "...", ...attributes}

Writing​

Writing is not supported.

Key Features​

  • Node + edge stream: uniform dict records with _type
  • Attribute passthrough: GraphML keys become record fields
  • NetworkX backed

Usage​

from iterable import open_iterable

with open_iterable("network.graphml") as source:
    for row in source:
        if row["_type"] == "node":
            print("node", row["id"])
        else:
            print("edge", row["source"], "->", row["target"])

Parameters​

No format-specific iterableargs.

Installation​

pip install 'iterabledata[graph]'

Requires networkx.

Limitations​

  1. Read-only
  2. Memory: entire graph loaded before iteration
  3. Requires networkx
  4. Hypergraphs / exotic GraphML extensions depend on NetworkX support

Error Handling​

  • ImportError: missing networkx — install iterabledata[graph]
  • Parse / I/O errors: invalid GraphML or missing files raise NetworkX or file exceptions
  • Format is registered writable=False
  • GEXF — Graph Exchange XML Format
  • DOT / GV — GraphViz DOT