Graph Visualization for Python by Neo4j

neo4j-viz is a Python package for creating interactive graph visualizations based on data from Neo4j products.

The output is of type IPython.display.HTML and can be viewed directly in a Jupyter Notebook, Streamlit. Alternatively, you can export the output to a file and view it in a web browser.

The package wraps the Neo4j Visualization JavaScript library (NVL).

Warning

This package is still in development and the API is subject to change.

Some notable features

• Easy to import graphs represented as:
  • projections in the Neo4j Graph Data Science (GDS) library
  • graphs from Neo4j query results
  • pandas DataFrames
• Node features:
  • Sizing
  • Colors
  • Captions
  • Pinning
  • On hover tooltip
• Relationship features:
  • Colors
  • Captions
  • On hover tooltip
• Graph features:
  • Zooming
  • Panning
  • Moving nodes
  • Using different layouts
• Additional convenience functionality for:
  • Resizing nodes, optionally including scale normalization
  • Coloring nodes based on a property
  • Toggle whether nodes should be pinned or not

Please note that this list is by no means exhaustive.

Getting started

Installation

Simply install with pip:

pip install neo4j-viz

Basic usage

We will use a small toy graph representing the purchase history of a few people and products.

We start by instantiating the Nodes and Relationships we want in our graph. The only mandatory fields for a node are the "id", and "source" and "target" for a relationship. But the other fields can optionally be used to customize the appearance of the nodes and relationships in the visualization.

Lastly we create a VisualizationGraph object with the nodes and relationships we created, and call its render method to display the graph.

from neo4j_viz import Node, Relationship, VisualizationGraph

nodes = [
    Node(id=0, size=10, caption="Person"),
    Node(id=1, size=10, caption="Product"),
    Node(id=2, size=20, caption="Product"),
    Node(id=3, size=10, caption="Person"),
    Node(id=4, size=10, caption="Product"),
]
relationships = [
    Relationship(
        source=0,
        target=1,
        caption="BUYS",
    ),
    Relationship(
        source=0,
        target=2,
        caption="BUYS",
    ),
    Relationship(
        source=3,
        target=2,
        caption="BUYS",
    ),
]

VG = VisualizationGraph(nodes=nodes, relationships=relationships)

VG.render()

This will return a IPython.display.HTML object that can be rendered in a Jupyter Notebook or streamlit application.

Please refer to the documentation for more details on the API and usage.

Examples

For some Jupyter Notebook and streamlit examples, checkout the /examples directory.

Contributing

If you would like to contribute to this project, please follow our Contributor Guidelines.