OPTICS Algorithm: Find clusters of related points using OPTICS

Recommend this page to a friend!
Stumble It! Stumble It! Bookmark in del.icio.us Bookmark in del.icio.us
  Info   View files View files (6)   DownloadInstall with Composer Download .zip   Reputation   Support forum (1)   Blog    
Last Updated Ratings Unique User Downloads Download Rankings  
2014-08-17 (1 year ago) RSS 2.0 feedNot enough user ratingsTotal: 123 All time: 329 This week: 21Up
Version License JavaScript version Categories  
optics 1.0BSD License3.0Graphics, Algorithms
Description Author  

This object can find clusters of related points in data sets using the OPTICS algorithm.

It takes a data set of points and can determine which points are related with each other using the OPTICS algorithm to group the points in clusters.

This object can render a bar chart in a canvas object with a representation of the clusters that were found, so it can be presented in a Web page.

The function to compute the distance between data set points can be redefined to handle data set points of any type.

Innovation Award  
JavaScript Programming Innovation award winner
July 2013
When you want to analyze large sets of sparse data it is hard to take conclusions about the most relevant groups of data you have.

This object provides an impressive solution to find clusters of 2D data points using the OPTICS algorithm.

Manuel Lemos
Picture of Franz Josef Brünner
  Performance   Level  
Name: Franz Josef Brünner <contact>
Classes: 6 packages by
Country: Austria Austria
Age: 30
All time rank: 311 in Austria Austria
Week rank: 15 Up1 in Austria Austria Equal
Innovation award
Innovation award
Nominee: 5x

Winner: 4x


This Project is made for a Bachelor-Thesis, so you may have to rewrite and modify some code before using OPTICS in you own Project.
The Clustering-Algorithm OPTICS is implemented purely in JavaScript as are the Files in this Project.
There are some Examples in this Project that can be used right now after downloading.

The HTML-File with HTML5-Webworker (optics_cluster_with_webworker.html) should be considered as a way to search for optimal Parameter-Values for OPTICS in a very short time, this File is also the aspect behind which this Project is made.

But there are other HTML-Files (optics_cluster_without_webworker.hmlt, parameter_selection.html) too that you can try out OPTICS manually with.

Hope this Project helps others in some way and most of all have fun when experimenting with OPTICS.
  Files folder image Files  
File Role Description
Files folder imagelibs (1 file, 1 directory)
Accessible without login Plain text file optics_cluster_with_webworker.html Example tool to try out
Accessible without login Plain text file parameter_selection.html Example tool to try out
Accessible without login Plain text file README.md Data readme of repo

  Files folder image Files  /  libs  
File Role Description
Files folder imagedatasets (2 files)
  Plain text file optics.js Class required classes for optics

  Files folder image Files  /  libs  /  datasets  
File Role Description
  Accessible without login Plain text file birth_and_death_rate.js Conf. a dataset that is used by optics
  Accessible without login Plain text file random_points.js Conf. a dataset that is used by optics

 Version Control Unique User Downloads Download Rankings  
 100%Total:123All time:329
 This week:0This week:21Up