Python implementation of k-means clustering algorithm with graphs created with matplotlib.

Run the program as "python kmeans.py". It opens and reads data from a file "input.txt" which is in the same directory. It shows this data on a 2D scatterplot to the user. Then it accepts number of clusters to create (k) and begins running the k-means algorithm. After it finishes execution, it shows a similar 2D plot but with points color coded based on the cluster they belong to. It will also show the corresponding cluster centres for each cluster - in its unique color.

As of now, the implementation is limited to 2 features in the input data i.e. 2 dimensional training sets.

The red points are each one training example represented by (X1_i, X2_i).

Each color i.e. Red, Blue, Green and Brown represents one cluster and the bold points represent their cluster centres.

• Python 2.7.x
• Library - matplotlib
• Library - numpy