https://github.com/sungwoncho/k_means_ruby

K-means Algorithm in Ruby
https://github.com/sungwoncho/k_means_ruby

Last synced: 3 months ago
JSON representation

K-means Algorithm in Ruby

• Host: GitHub
• URL: https://github.com/sungwoncho/k_means_ruby
• Owner: sungwoncho
• License: mit
• Created: 2015-03-13T11:37:37.000Z (almost 10 years ago)
• Default Branch: master
• Last Pushed: 2015-03-14T02:37:57.000Z (almost 10 years ago)
• Last Synced: 2024-10-09T09:44:02.407Z (3 months ago)
• Language: Ruby
• Size: 148 KB
• Stars: 10
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: MIT-LICENSE

Awesome Lists containing this project

README

        
# k_means_ruby

[![Build Status](https://travis-ci.org/sungwoncho/k_means_ruby.svg?branch=master)](https://travis-ci.org/sungwoncho/k_means_ruby)

[![Coverage Status](https://coveralls.io/repos/sungwoncho/k_means_ruby/badge.svg?branch=master)](https://coveralls.io/r/sungwoncho/k_means_ruby?branch=master)

An implementation of k-means algorithm in Ruby.

## Definition

K-means algorithm is a clustering algorithm for a Euclidean space. It assumes that there are k clusters in the beginning.

* Choose k points that are likely to be in different clusters.

* Make those the points centroids of clusters.

* For the remaining points:

  * Find the centroid to which the point is closest.

  * Merge the point into that cluster.

  * Recalculate the centroid of that cluster.

## Use

In your ruby console, require the main file:

`require './lib/k_means_ruby.rb'`

Define `Points` and `Clusters`:

```

p1 = KMeansRuby::Point.new(1,2)

p2 = KMeansRuby::Point.new(10,19)

...

c1 = KMeansRuby::Cluster.new(p1)

...

```

Instantiate the `Algorithm`:

`algorithm = KMeansRuby::Algorithm.new(points=[p1, p2], clusters=[c1])`

Run the algorithm by calling `algorithm.run_once` or `algorithm.repeat_for(n)`

## Contributing

### TODO:

* Allow users to run the algorithm until the clusters do not change anymore.