Customized K-Means

My implementation of K means algorithm is customized for different need. Initial cluster centroid can be selected in various of ways. Those are:
• Randomly initialized cluster centroid as one of the data row.
• Select first 3 data row was the three cluster center.
• Provide the cluster centroid as a parameter, it is specially helpful when you want to perform the cluster with the same initial data centers so that we don’t have to worry about K means naming different to the same cluster in different run.