VivaGraphJS is a free graph drawing library for JavaScript. It is designed to be extensible and to support different rendering engines and layout algorithms. At the moment it supports rendering graphs using WebGL, SVG or CSS formats. Layout algorithms currently implemented are:
- Force Directed - based on Barnes-Hut
simulation and optimized for JavaScript language this algorithm gives
N * lg(N) + V
performance per iteration.
Library provides API which tracks graph changes and reflect changes on the rendering surface accordingly.
Some examples of library usage in the real projects:
- Amazon Visualization Shows related products on Amazon.com, uses SVG as graph output
- YouTube Visualization Shows related videos from YouTube. SVG based.
- Facebook Visualization friendship visualization on Facebook. WebGL based.
- Graph Viewer visualization of sparse matrices collection of the University of Florida. WebGL based.
- Vkontakte Visualization friendship visualization of the largest social network in Russia vk.com. WebGL based.
To start using the library include vivagraph.js
script from the dist folder. The following code is the minimum required to render a graph with two nodes and one edge:
var graph = Viva.Graph.graph();
graph.addLink(1, 2);
var renderer = Viva.Graph.View.renderer(graph);
renderer.run();
This will produce the following layout:
The code above adds a link to the graph between nodes 1
and 2
. Since nodes are not yet in the graph
they will be created. It's equivalent to
var graph = Viva.Graph.graph();
graph.addNode(1);
graph.addNode(2);
graph.addLink(1, 2);
var renderer = Viva.Graph.View.renderer(graph);
renderer.run();
VivaGraphJS is all about customization. You can easily change the appearance of nodes and links. You can also change the layouting algorithm and medium that displays elements of the graph. For example: The following code allows you to use CSS-based rendering, instead of the default SVG.
var graph = Viva.Graph.graph();
graph.addLink(1, 2);
var graphics = Viva.Graph.View.cssGraphics();
var renderer = Viva.Graph.View.renderer(graph,
{
graphics : graphics
});
renderer.run();
graphics
class is responsible for rendering nodes and links on the page. And renderer
orchestrates the process. To change nodes appearance tell graphics
how to represent them. Here is an example of graph with six people who I follow at github:
var graph = Viva.Graph.graph();
// Construct the graph
graph.addNode('anvaka', {url : 'https://secure.gravatar.com/avatar/91bad8ceeec43ae303790f8fe238164b'});
graph.addNode('manunt', {url : 'https://secure.gravatar.com/avatar/c81bfc2cf23958504617dd4fada3afa8'});
graph.addNode('thlorenz', {url : 'https://secure.gravatar.com/avatar/1c9054d6242bffd5fd25ec652a2b79cc'});
graph.addNode('bling', {url : 'https://secure.gravatar.com/avatar/24a5b6e62e9a486743a71e0a0a4f71af'});
graph.addNode('diyan', {url : 'https://secure.gravatar.com/avatar/01bce7702975191fdc402565bd1045a8?'});
graph.addNode('pocheptsov', {url : 'https://secure.gravatar.com/avatar/13da974fc9716b42f5d62e3c8056c718'});
graph.addNode('dimapasko', {url : 'https://secure.gravatar.com/avatar/8e587a4232502a9f1ca14e2810e3c3dd'});
graph.addLink('anvaka', 'manunt');
graph.addLink('anvaka', 'thlorenz');
graph.addLink('anvaka', 'bling');
graph.addLink('anvaka', 'diyan');
graph.addLink('anvaka', 'pocheptsov');
graph.addLink('anvaka', 'dimapasko');
// Set custom nodes appearance
var graphics = Viva.Graph.View.svgGraphics();
graphics.node(function(node) {
// The function is called every time renderer needs a ui to display node
return Viva.Graph.svg('image')
.attr('width', 24)
.attr('height', 24)
.link(node.data.url); // node.data holds custom object passed to graph.addNode();
})
.placeNode(function(nodeUI, pos){
// Shift image to let links go to the center:
nodeUI.attr('x', pos.x - 12).attr('y', pos.y - 12);
});
var renderer = Viva.Graph.View.renderer(graph,
{
graphics : graphics
});
renderer.run();
The result is:
Graphs vary by their nature. Some graphs have hundreds of nodes and few edges (or links), some might connect every node with every other. Tuning the physics often helps get the best layout. Consider the following example:
var graphGenerator = Viva.Graph.generator();
var graph = graphGenerator.grid(3, 3);
var renderer = Viva.Graph.View.renderer(graph);
renderer.run();
Graph generators are part of the library, which can produce classic graphs. grid
generator creates a grid with given number of columns and rows. But with default parameters the rendering is pretty ugly:
Let's tweak the original code:
var graphGenerator = Viva.Graph.generator();
var graph = graphGenerator.grid(3, 3);
var layout = Viva.Graph.Layout.forceDirected(graph, {
springLength : 10,
springCoeff : 0.0005,
dragCoeff : 0.02,
gravity : -1.2
});
var renderer = Viva.Graph.View.renderer(graph, {
layout : layout
});
renderer.run();
Now the result is much better:
Tuning layout algorithm is definitely one of the hardest part of using this library. It has to be improved in future to simplify usage. Each of the force directed algorithm parameters are described in the source code.
Run the following script:
git clone git://github.com/anvaka/VivaGraphJS.git
cd ./VivaGraphJS
npm install
grunt
The combined/minified code should be stored in dist
folder.
Please refer the upgrade guide to see how to update older versions of the library to the latest one.
Disclaimer: I wrote this library to learn JavaScript. By no means I pretend to be an expert in the language and chosen approach to design may not be the optimal. I would love to hear your feedback and suggestions.
Though I implemented this library from scratch, I went through many existing libraries to pick the best (at my view) out of them. If you are evaluating libraries for your project make sure to check them out as well:
- Dracula Graph Library - written by Johann Philipp Strathausen and uses Raphaël library to render graphs. Has very simple API.
- D3 - one of the best data visualization library in JavaScript world. From Mike Bostock.
My goal is to create highly performant javascript library, which serves in the field of graph drawing. To certain extent I achieved it. But I have no doubt there is much more to improve here.