Skip to content

uhho/density-clustering

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

27 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Density Based Clustering for JavaScript

Package contains popular methods for cluster analysis in data mining:

  • DBSCAN
  • OPTICS
  • K-MEANS

Overview

DBSCAN

Density-based spatial clustering of applications with noise (DBSCAN) is one of the most popular algorithm for clustering data.

http://en.wikipedia.org/wiki/DBSCAN

OPTICS

Ordering points to identify the clustering structure (OPTICS) is an algorithm for clustering data similar to DBSCAN. The main difference between OPTICS and DBSCAN is that it can handle data of varying densities.

http://en.wikipedia.org/wiki/OPTICS_algorithm

Important

Clustering returned by OPTICS is nearly indistinguishable from a clustering created by DBSCAN. To extract different density-based clustering as well as hierarchical structure you need to analyse reachability plot generated by OPTICS.

For more information visit http://en.wikipedia.org/wiki/OPTICS_algorithm#Extracting_the_clusters

K-MEANS

K-means clustering is one of the most popular method of vector quantization, originally from signal processing. Although this method is not density-based, it's included in the library for completeness.

http://en.wikipedia.org/wiki/K-means_clustering

Installation

Node:

npm install density-clustering

Browser:

bower install density-clustering
# build
npm install
gulp

Examples

DBSCAN

var dataset = [
    [1,1],[0,1],[1,0],
    [10,10],[10,13],[13,13],
    [54,54],[55,55],[89,89],[57,55]
];

var clustering = require('density-clustering');
var dbscan = new clustering.DBSCAN();
// parameters: 5 - neighborhood radius, 2 - number of points in neighborhood to form a cluster
var clusters = dbscan.run(dataset, 5, 2);
console.log(clusters, dbscan.noise);

/*
RESULT:
[
    [0,1,2],
    [3,4,5],
    [6,7,9],
    [8]
]

NOISE: [ 8 ]
*/

OPTICS

// REGULAR DENSITY
var dataset = [
  [1,1],[0,1],[1,0],
  [10,10],[10,11],[11,10],
  [50,50],[51,50],[50,51],
  [100,100]
];

var clustering = require('density-clustering');
var optics = new clustering.OPTICS();
// parameters: 2 - neighborhood radius, 2 - number of points in neighborhood to form a cluster
var clusters = optics.run(dataset, 2, 2);
var plot = optics.getReachabilityPlot();
console.log(clusters, plot);

/*
RESULT:
[
  [0,1,2],
  [3,4,5],
  [6,7,8],
  [9]
]
*/
// VARYING DENSITY
var dataset = [
  [0,0],[6,0],[-1,0],[0,1],[0,-1],
  [45,45],[45.1,45.2],[45.1,45.3],[45.8,45.5],[45.2,45.3],
  [50,50],[56,50],[50,52],[50,55],[50,51]
];

var clustering = require('density-clustering');
var optics = new clustering.OPTICS();
// parameters: 6 - neighborhood radius, 2 - number of points in neighborhood to form a cluster
var clusters = optics.run(dataset, 6, 2);
var plot = optics.getReachabilityPlot();
console.log(clusters, plot);

/*
RESULT:
[
  [0, 2, 3, 4],
  [1],
  [5, 6, 7, 9, 8],
  [10, 14, 12, 13],
  [11]
]
*/

K-MEANS

var dataset = [
  [1,1],[0,1],[1,0],
  [10,10],[10,13],[13,13],
  [54,54],[55,55],[89,89],[57,55]
];

var clustering = require('density-clustering');
var kmeans = new clustering.KMEANS();
// parameters: 3 - number of clusters
var clusters = kmeans.run(dataset, 3);
console.log(clusters);

/*
RESULT:
[
  [0,1,2,3,4,5],
  [6,7,9],
  [8]
]
*/

Testing

Open folder and run:

mocha -R spec

License

Software is licensed under MIT license. For more information check LICENSE file.

About

Density Based Clustering in JavaScript

Resources

License

Stars

Watchers

Forks

Packages

No packages published