Calculating Distances Between Points and Centroids in K-Means Clustering: A Workaround for Single-Centroid Clusters