[Numpy-discussion] Scipy 0.6.0 to 0.7.0, sparse matrix change

Nathan Bell wnbell@gmail....
Wed Jun 17 01:02:38 CDT 2009


On Mon, Jun 15, 2009 at 6:47 AM, Fadhley Salim
<Fadhley.Salim@uk.calyon.com> wrote:
>
> I'm trying to track down a numerical discrepancy in our proejct. We
> noticed that a certain set of results are different having upgraded from
> scipy 0.6.0 to 0.7.0.
>
> The following item from the Scipy change-log is our current number-one
> suspect. Could anybody who knows suggest what was actually involved in
> the change which I have highlighted with stars below?
>
> [...]
>
> The handling of diagonals in the ``spdiags`` function has been changed.
> It now agrees with the MATLAB(TM) function of the same name.
>
> *** Numerous efficiency improvements to format conversions and sparse
> matrix arithmetic have been made.  Finally, this release contains
> numerous bugfixes. ***
>

Can you elaborate on why you think sparse matrices may be the culprit?
 None of the changes between 0.6 and 0.7 should produce different
numerical results (beyond standard floating point margins).

--
Nathan Bell wnbell@gmail.com
http://www.wnbell.com/


More information about the Numpy-discussion mailing list