mottalrd / k_means

A simple K means implementation

Installation

This code has been tested with Python 2.7.11

First create a virtual environment

$: pyenv versions
  system
  2.7.11rc1
  3.5.0
$: pyenv virtualenv 2.7.11rc1 k_means
$: pip install -r requirements.txt

You can now load your KMeans code in a console:

>>> import imp  
>>> k_means = imp.load_source('k_means', 'lib/k_means.py')  
>>> data = np.array([  
  [1, 1],  
  [2, 2],  
  [3, 3],  
  [10, 10],  
  [11, 11],  
  [12, 12]  
])  
>>> k_means.KMeans(data, 2).fit()  
[Cluster(center=[ 11.  11.], members=[array([10, 10]), array([11, 11]), array([12, 12])], converged=True), Cluster(center=[ 2.  2.], members=[array([1, 1]), array([2, 2]), array([3, 3])], converged=True)]

Alternatively you can download the tar in the dist folder, extract it and install it with

$: python setup.py install

Finally, check the experiments folder to see K means in action on the iris dataset in a Jupyter notebook.

Running the tests

py.test