Extend your graphs beyond structure—add meaning with ugraph.
ugraph builds on igraph to provide a powerful way to define and work with custom node and link types in your graphs. This package is ideal for those who need more than just graph structure—it empowers you to combine graph data with rich information via Python dataclasses, and comes with built-in support for JSON storage and 3D visualizations.
Because your graphs aren't just for you
(igraph → ugraph)
Graphs often represent more than their edges and nodes—they carry data, behaviors, and relationships that need to be understood in context. ugraph bridges this gap by enabling:
- Custom node and link classes: Add type-safe attributes and behaviors to your graph elements.
- Data serialization: Easily save and load your graphs in JSON format for persistence and sharing.
- 3D visualization: Render interactive, browser-based 3D visualizations in HTML using Plotly.
With ugraph, your graphs are as understandable and maintainable as the data they represent.
- Custom Classes: Define your nodes and links as Python dataclasses, allowing for type hints, IDE autocompletion, and type checking.
- Serialization: Store and reload your networks seamlessly using JSON files.
- Interactive Visualization: Generate 3D plots of your graphs in HTML for better insights and presentation.
ugraph is not intended for creating graph figures or visualizations (e.g., bar charts, scatter plots). It is a tool for working with graph data structures (nodes and links) and enhancing their usability.
Install ugraph using pip:
pip install ugraph(if you want plotting included, use pip install ugraph[plotting])
ugraph works similarly to igraph, with the added flexibility of custom node and link types. You can define attributes like coordinates, IDs, or any other domain-specific data in a type-safe and Pythonic way.
Explore usage examples in the examples directory, or start with a minimal example.
This project builds upon the excellent igraph library. We acknowledge and thank the igraph community for their foundational work.
See the LICENSE file for rights and limitations (MIT).