Being a newbie in R, I'm not very sure how to choose the best number of clusters to do a k-means analysis
