Le 11/08/2010 16:00, Dimitri Pourbaix a écrit :
> Hi,
> 
> While fixing MATH-405, I noticed that some unit tests were failing with
> the corrected version!  Obviously, some results against which those tests
> are competing were obtained ... with the bugged code.  The usefulness of
> such tests is therefore pretty limited.

>From what I recall, most of the results were obtained from the fortran
Netlib library. However, some results may have been produced by our own
code and only checked for consistency. This is weird and should be
avoided as much as possible, of course. Verifying all tests would be a
daunting task as we have more than 2000 tests now, so I guess we will
find some cases like this as we update some code.

> 
> It also appears that the Cartesian norm of a vector (in AbstractRealVector,
> in AbstractLeastSquares, ...) does not care at all about overflow or
> underflow.  I thus translated enorm.f (minpack) into Java.  My initial
> intend was to put in AbstractLeastSquares but it might be more useful in
> a more general location (MathUitls).  What do you think?

Very good idea!
Perhaps this could be put in MatrixUtils, as this class already holds
some utility methods for vectors in addition to utility methods for
matrices.

thanks a lot
Luc

> 
> Regards,
>  Dim.
> ----------------------------------------------------------------------------
> 
> Dimitri Pourbaix                         *
> Institut d'Astronomie et d'Astrophysique *      Don't worry, be happy
> CP 226, office 2.N4.211, building NO     *         and CARPE DIEM.
> Universite Libre de Bruxelles            *
> Boulevard du Triomphe                    *      Tel : +32-2-650.35.71
>  B-1050 Bruxelles                        *      Fax : +32-2-650.42.26
> http://sb9.astro.ulb.ac.be/~pourbaix     * mailto:pourb...@astro.ulb.ac.be
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
> For additional commands, e-mail: dev-h...@commons.apache.org
> 
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org

Reply via email to