pvsvg

Introduction

a python wrapper of the vis.js network visualization library with an additional ability to export static SVG.

Inspired by the excellent python library pyvis but I required a function to export the resulting network as a static SVG.

The code for exporting the SVG is taken from justinharrell/vis-svg.

Installation

You can install this via pip:

pip install pvsvg

Usage

This library exposes a single class ( Network ) which takes as input a networkx Graph.

All the configuration for node and edge level attributes (like color, size, mass) are passed transparently to vis.js so you can configure your resulting network via the attributes of your networkx nodes and edges respectively.

You can see node level configuration and edge level configuration on the vis.js documentation.

Exporting as SVG

The resulting HTML file will have some configuration buttons built-in. One of these is the export-svg button which will generate the static SVG and write it to a file (network.svg)

Keep in mind that it will write the SVG of whatever is within the Canvas element at the time. Essentially the view you have within the boundaries of the Canvas is exactly what will be written to the SVG.

Note: Currently the node borders are not being written to the SVG. Besides that everything else should appear.

Example

from pvsvg import Network
import networkx as nx

# generate some random graph
g = nx.fast_gnp_random_graph(15, 0.2)

# set node level attributes
for n in g.nodes(data=True):
    id = n[0]
    if id % 2 == 0:
        n[1]["color"] = "red"
    else:
        n[1]["color"] = "blue"

    if id % 3 == 0:
        n[1]["shape"] = "square"
    else:
        n[1]["shape"] = "dot"

    n[1]["label"] = f"Node {id}"
    n[1]["title"] = f"Node {id} has color {n[1]['color']} and shape {n[1]['shape']}"
    n[1]["size"] = (id + 1)

# set edge level attributes
for e in g.edges(data=True):
    e[2]["color"] = {
        "color": "black",
        "inherit": False,
    }

# Instantiate the network
n = Network(g, width=1200)

# Draw the network as an html
n.draw("test.html")