I am porting python code to Sho, and I am having a trouble indexing the matrix as in python like this:
(D, W) = scipy.linalg.eig(A)
W = W[:, D>0]
This matrix is too huge to be copied to a new array in the memory, so I was looking for an optimized implementation.
Thanks in advance.