magnitude of column vectors in a matrix

[michael_mc]

Hello all,

I'm new to LAPACK and I am trying to compute the magnitude (square of the l2 norm ) of each of a set of vectors stored in a matrix (one vector per column).

For example, if the dimensionality of the vectors is 10, and if I have 200 vectors, the matrix will be 200x10 .
I need to output a 1x200 matrix where each column is the magnitude of each of the 200 vectors.

I know that there is a routine in the BLAS leve-1 that can compute the L2 norm of only one vector. In my case, I need to efficiently do the computation for millions of vectors.


PS: This is how I do this in MATLAB: result= sum(v.^2); where v is a matrix with one vector per column

Thank you for your attention,
best rgs,
Michael C.