Quickstart

nxv renders NetworkX graphs using GraphViz.

• Using nxv inside of Jupyter is the easiest and recommended way to get started.

• Using nxv outside of Jupyter describes how to use nxv in other settings.

Using nxv inside of Jupyter

Start by importing networkx and nxv.

import networkx as nx
import nxv

Define a simple NetworkX graph.

graph = nx.Graph()
graph.add_edge("A", "B")
graph.add_edge("B", "C")
graph.add_edge("C", "D")
graph.add_edge("B", "E")

Render the graph with GraphViz using the render() function.

nxv.render(graph)

Use a Style to specify how to style the graph using GraphViz.

style = nxv.Style(
    graph={"rankdir": "LR"},
    node={"shape": "square"},
    edge={"style": "dashed"},
)

See the GraphViz attributes documentation for information on what attributes are available to use.

Render the graph with the Style by passing it to the render() function.

nxv.render(graph, style)

The Style parameters can be functions that map the parts of a graph to different styles.

style = nxv.Style(
    graph={"rankdir": "LR"},
    node=lambda u, d: {"shape": "circle" if u in "AEIOU" else "square"},
    edge=lambda u, v, d: {"style": "dashed", "label": u + v},
)

nxv.render(graph, style)

Using nxv outside of Jupyter

Outside of Jupyter, the format parameter of the render() function is required. When the format parameter is provided, the behavior of the render() function is to return the bytes of the result in the specified format.

data = nxv.render(graph, style, format="svg")
with open("graph.svg", "wb") as f:
    f.write(data)