So The concept is If you have a minimal variety of types with your data, then you should just rely how frequently Just about every classification occurs.The location exactly where we're going to see the dissimilarities from the data documents, I'll shift more than on the data file.And so During this photo, we have been likely from a sort of a three