On Feb 1, 9:21 pm, John H Palmieri <jhpalmier...@gmail.com> wrote:
> On Feb 1, 4:44 pm, mabshoff <mabsh...@googlemail.com> wrote:
<SNIP>
> > Can you default it to LinBox in your case and report if LinBox causes
> > trouble either by throwing an error or by producing wrong results?
>
> I don't think I can. Running mat.elementary_divisors
> (algorithm='linbox') raises an exception, saying how broken linbox
> is. If the exception weren't there, the code seems to want to call
>
> mat._elementary_divisors_linbox()
>
> which gives me a NotImplementedError. Oh, well, of course it does:
> look at the code for smithform in libs/linbox/linbox.pyx. I don't know
> Cython, and if I comment out the "raise NotImplementedError" line and
> uncomment everything else, I get errors when restarting sage with
> 'sage -br':
>
> ImportError: dlopen(/Applications/sage_builds/sage-3.3.alpha1/local/
> lib/python2.5/site-packages/sage/libs/linbox/linbox.so, 2): Symbol not
> found: _linbox_integer_dense_smithform
Hmm, we might need to add back support for this in the Sage wrapper of
LinBox itself.
> Should I open a trac ticket about fixing the linbox implementation of
> Smith normal form? Searching trac for "linbox smith" and "linbox
> elementary" doesn't lead to anything relevant, and neither does
> browsing the open tickets in the linear algebra section.
I would suggest we ask Clement what the status if the Smith normal
form in LinBox is. If it is supposed to be stable and working we can
open a ticket to get it back into the wrapper and then also use it
from Sage via another ticket.
Since Clement is rather busy the last couple months it might be a
while until we hear from him. I would suggest someone asks about this
in the linbox-use Google group.
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
