Fruchterman Reingold

Implementation of the Fruchterman Reingold algorithm for Javascript and graphology. This implementation is compatible to @ambalytics/webgraph.

This implementation is based on the original algorithm published in 1991 (referenced below). I made several small additions:

• While calculating the forces of a node the implementation skips nodes that are far away (1000 * k). These nodes have a small (negligible) influence on the force.
• The attractive force are scaled by the link weight.
• The implementation supports a gravity forced directed to the center.
• The forces can be scaled by a speed setting.

(Btw: inspired by graphology-layout-forceatlas2)

Reference

Fruchterman, T.M.J. and Reingold, E.M. (1991), Graph drawing by force-directed placement. Softw: Pract. Exper., 21: 1129-1164. https://doi.org/10.1002/spe.4380211102

Installation

As this package hosted is on github packages, you have to login to this registry.
You can find all details here and about the authentication here.
Basically you need to generate a PAT with at least repo and read:packages rights. Afterwards you have to call the login command stated below and enter your username, token as password and email when prompted.

npm login --scope=@ambalytics --registry=https://npm.pkg.github.com

npm install @ambalytics/graphology-layout-fruchtermanreingold

Usage

Settings

parameter	type	default	decription
iterations	number	10	amount of iterations of the algorithm
edgeWeightInfluence	number	1	influence of the edge's weights on the layout
speed	number	1	scaling factor for forces
gravity	number	10	strength of the layout's gravity
C	number	1	C factor (scales the k coefficient)
weightAttribute	string	weight	edge attribute to use as weight

Synchronous layout

import fruchtermanReingold from '@ambalytics/graphology-layout-fruchtermanreingold';

const positions = fruchtermanReingold(graph);

// With settings:
const positions = fruchtermanReingold(graph, {
  iterations: 50,
  edgeWeightInfluence: 5,
});

// To directly assign the positions to the nodes:
fruchtermanReingold.assign(graph);

Arguments

parameter	type	decription
graph	Graph	amount of iterations of the algorithm
options?	Partial<FruchtermanReingoldLayoutOptions>	influence of the edge's weights on the layout

Webworker

This library also supports the computation with a webworker. The syntax for that is nearly the same. In this case the return values will be wrapped in a Promise. To use the webworker you have to change the import statement to:

import fruchtermanReingold from '@ambalytics/graphology-layout-fruchtermanreingold/worker';