Bluemi/clustering-testcase

Implementation of some clustering algorithms.

clustering-testcase

Implementation of some clustering algorithms.

Controls

• c: Toggles visualization of clustering algorithm
• + / -: Use more or less cluster centroids
• a / A: Use next/prev cluster algorithm
• n / N: Use next/prev image/distribution
• 2: Show 2d point distribution
• 3: Show images. Colors are interpreted as 3d points and clustered
• r: resample 2d distribution
• Esc: Quit the program

Credits

• The generalized Lloyd Algorithm is taken from spencerkent, but I did not manage to get it running.
• The K-Means Clustering uses the scipy implementation.