I have used the dgesvd routine in CLAPACK to calculate the best-fit plane parameters for a point-cloud. The input matrix M is the Nx4 matrix containing the data points and the last column of ones. I am doing the SVD of the matrix M to find the singular vectors U and V as M = USV'. From there. I am finding the singular vector corresponding to the smallest singular values. Now I also want to find the covariance of the best-fit plane from the U,S and V matrices found. Is there any easy method to do so?