coresets.py - This program generates a lightweight coreset using the algorithm from Bachem, Lucic, Krause "Scalable k-Means Clustering via Lightweight Coresets"
To run this code, python version must be greater than Python 3.6
usage: python coresets.py dataset_filename m dataset_filename: filename of dataset to be converted m: number of subsamples in lightweight coreset example: python coresets.py bio_train.dat 1000
In the experiments done in the paper, the values of m used were: {1000, 2000, 5000, 10000, 20000}
Once the lightweight coresets are computed, they are exported to a file called "export.dat" in the same directory. The resulting coresets are evaluated in a separate k-means++ algorithm and compared for accuracy and performance