SpatialGraphLib.js

v0.10

This library provides tools to generate and examine spatial graphs, i.e. graphs that are embedded into space. It includes a reference implementation of the Scale-Invariant Spatial Graph (SISG) Model. In addition to spatial graphs, the generation of some non-spatial graph models is supported.

Content

• Literature
• Demo
• Usage
• License

Literature

The functionality is, in parts, based on research conducted by the author. In particular, the definition of the Scale-Invariant Spatial Graph Model is contained in the following paper:

Franz-Benjamin Mocnik, Andrew U. Frank: Modelling Spatial Structures
Proceedings of the 12th Conference on Spatial Information Theory (COSIT), 2015

You find more information on http://www.mocnik-science.net

Demo

A demo of the functionality can be found at: http://projects.mocnik-science.net/spatial-graphs

Usage

The library can be used online in a browser or by node.js.

A new graph can be created as an instance of Graph:

var g = new Graph();

This graph can be modified by executing operations on the graph. These operations return the graph, allowing to execute some operations in a row. For example, we can write:

g.addNodesRandom(100, [51, 52], [0, 8]).addEdgesSISGModel(1.5);

Basic data of the graph

The functions getNodes and getEdges return the nodes resp. the edges of the graph:

g.getNodes();
g.getEdges();

Basic modification of the graph

You can add nodes to the graph by using the function addNodes:

g.addNodes([{coordinates: [52, 8]}, {coordinates: [51, 0]}])

The nodes are assumed to be objects containing the key coordinates with associated value a list consisting of the values for the latitude and longitude.

The edges can be removed from the graph by writing:

g.clearEdges();

The nodes and edges can be removed by writing:

g.clearGraph();

You can ensure that the nodes and edges are unique by executing the functions nubNodes and nubEdges; double nodes and edges are removed.

g.nubNodes();
g.nubEdges();

Many edge generation algorithms of this library remove double edges automatically.

You can add n random nodes with latitude and longitude in given intervals like follows:

g.addNodesRandom(n, latInterval, lonInterval);

The intervals are of the form [a, b].

Add nodes from file

You can load points from a GeoJSON file into the graph:

g.addNodesGeoJSON(JSON.parse(...));

In a similar way, you can load data from a CSV file into the graph:

g.addNodesGeoCSV(...);

The first column of the file is assumed to contain the latitude, the second one the longitude. If a third column is provided, it is assumed to contain a label for the node.

Generation of edges: Scale-Invariant Spatial Graph (SISG) Model

Edges can be added to a graph according to the SISG model (see literature) using the function addEdgesSISGModel. It assumes that the graph already contains a number of nodes. The density parameter k is given as the only parameter:

g.addEdgesSISGModel(k);

Generation of edges: Gilbert model

Edges can be added according to the Gilbert model with given probability p as follows:

g.addEdgesGilbert(p);

Export

The graph can be exported in the CSV format (comma-separated values) as well as in the TGF format (trivial graph format). A string containing the data is returned.

g.csv();
g.tgf();

License

Copyright (c) 2015 by Franz-Benjamin Mocnik
The library is free to use for non-commercial projects.
Copys of the library must always contain this license.