Hi Eddy,
>1) Did you introduce long double because your
> algorithms need a higher accuracy ?
>
I introduced them because I demanded highest accuracy from the start,
not because I had proof Double_t wasn't enough.
> If so,
> a. That is extremely worrisome, have a look at
> "Accuracy and Stability of Numerical
>Algorithms"
> by Nicholas Higham
> b. Actually do not , there is money to be made
> from your algorithms :)
>
Not if I can never finish. In that case all I will reap is
unemployment. But the Windows results are *really* different and
non-sensical, so I can't just push ahead.
>2) Are you seeing differences between Windows and
> linux using TMatrix . That should not be the case.
>
I made my 'port to Windows, omitting the TMatrixQ because of statments
like yours. Now I see differences. All the rest of the code is as
close to the same as possible. Once I finish implementing I will see if
the results are then the same. If so, that will be the example you
want, but it's proprietary and I can't provide it. If it is still not
the same, then I'll look elsewhere.
In the event I can finger the accuracy as a source of my problem, I'm
sure I can put together a public example that highlights the problem.
But not today.
- J
--
Dr. John Krane
jkrane@netzero.com
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:07 MET