[Numpy-discussion] Change of SVD output

Robert Kern robert.kern@gmail....
Fri Jul 25 17:10:55 CDT 2008


On Fri, Jul 25, 2008 at 17:01, Arnar Flatberg <arnar.flatberg@gmail.com> wrote:
> Hi
>
> In a recent thread there was an error in how a matrix is reconstructed from
> its SVD decomposition. I apologize if this is just an old and settled issue
> and I am just adding noise, but I got bitten by numpy's unfamiliar output
> myself a long time ago and I see others get confused as well. So what is the
> issue?
>
> Let the svd decomposition of X be USV:
> U, S, V = np.linalg.svd(X),
> Then U has X's left singular vectors in its columns, S contains the singular
> values, and V has X's right singular vectors in its *rows*. The
> reconstruction of X is (matrix notation) will be: U*diag(S)*V  (not
> U*diag(S)*V.T).
> All other high-level software (Matlab, Scilab, Mathematica, R, etc), outputs
> the right singular vectors columnwise, that is V = [v1, v2, v3, ... vn],
> where vn is a column (eigenvector) in V, thus the reconstruction would be
> U*diag(S)*V.T. Also, as far as I know, most linear algebra textbooks operate
> with eigenvectors consistent as column vectors in explanations of the SVD. I
> think numpy's svd should do so too.
>
> I know lapack's dgesdd returns V.T (or conjugate), and specify that in its
> documentation so this is a true interface of the library, but I still think
> its wrong and its just too confusing for any beginner who usually has
> experience in other software, such as matlab, prior to numpy. Also, ,as I am
> typing here, I realize that changing the output would break lots of stuff,
> and pass silently through many tests as the shape of V is similar (if
> full_matrices=0). Oh well, I guess that proposal is off the table?

Yes.

> Perhaps
> some *stronger* hints in the documentation are needed.

Please do.

  http://sd-2116.dedibox.fr/pydocweb/wiki/Front%20Page/

> Arnar
>
> PS:
> In the docs at http://www.scipy.org/NumPy_for_Matlab_Users , the svd
> equivalents have wrong notation, this is not helping :-). I didnt manage to
> change it, perhaps some other may be so kind?

You will need to register an account. Make an account here:

  http://www.scipy.org/UserPreferences

If you do have an account, and you are still having problems, it may
be because of the account filtering I am using to try to control the
spam problem. Try a different account name.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
 -- Umberto Eco


More information about the Numpy-discussion mailing list