Python

Generate some data using:

Deterministic data generation algorithm (you can make your own)

Non deterministic data generation - again you can make your own if you like.

Make some data, using each algorithm, centered about the points (5, 5), (-5, 5), (-5, -5), and (5, -5). Make your data radius to be 3, 5, and 7. Show plots of the data for each algorithm. Try 3 different population sizes of the clusters, sparse like 10, medium 30, and lots 100. At the larger radius sizes the points from the clusters should overlap. You should save the data to a file, or make it so that your program will generate the same data each time by specifying the random seed.

Now implement some clustering algorithms:

Try doing the ad-hoc routine, one like we did in class, or your modded version of it.

Implement K-means.

Run these algorithms on your generated data sets. Make graphs that indicate somehow what clusters the points are in.

For the K-means algorithm, make sure that you can:

Do the basic algorithm some number of iterations

Mod the basic algorithm to stop when K centroids stop moving

Mod the algorithm so that a large value of K can be entered and the like centroids can be merged after the algorithm has found equilibrium.

Shows all parts of the assignment with the graphs, code, and a discussion of each part