I have a matrix M of m*n dimension. M contains n number of data each has m dimension and m is very very large than n.
Now my question is, how to compute or what are the steps or procedure to find PCA of M using SVD in OpenCV keeping only those eigenvectors containing 99% of total load or energy ?