The method works as follows. 1. For a data set with dimensionality, d, compute the variance of data in each dimension (column). 2. Find the column with maximum variance call it cvmax and sort it in any order. 3. Divide the data points of cv