Features

Algorithms

  • hard-kmeans clustering
  • soft-kmeans clustering
  • mixture-models clustering

Command line interface

  • standards compliant command line interface
  • control
    • which data to operate on
    • how many clusters to find
    • which algorithm and parameters to use

GUI

  • 3D-visualization using OpenGL
  • Zooming
  • Rotating (without the problem of the gimbal lock - we use quaternion math here)
  • display usage information
  • after the clustering process has finished you can step through the history of the process using "b" to go back one step in time, and "n" to advance to the next.
  • show/hide the data points, the cluster center trajectories, the cluster center representations.
  • the clustering process can be reinitialized and restarted without quitting the program.