siddheshk / faster-kmeans Goto Github PK

While result of SKlearn's K_means_elkan is all the same however much times you try.
So it's what caused the result of kpp.py+triangleInequality.py changable.

Hi,

included these lines in your code heuristicenhanced_kmeans.py:
pointList = loadDataset("Dataset/covtype.data/covtype.data",",")
Kmeans(16, pointList, 0.01, 20, initialCentroids=None)

Should i add any other function(may be main) to get the required results?

Also what are the suggested values of k, kmeansThreshold, centroidsToRemember

Hi~
I unzipped the *.zip files and got several kinds of files (some with txt extension, and others no extension at all). Would you please tell me how to read these files into pointList that the algorithm needs. Thank you!