Emmanuel Thomé wrote:
> On 24 juin, 07:50, Jason Grout <[email protected]> wrote:
>> And 5. Compare (maybe wrap??) to the nice LU decomposition routine that
>> already exists in sympy!  I just found this.  It looks like very nice
>> work.  I should probably look at sympy more often when there are missing
>> bits in Sage!
> 
> For symbolic matrices, comparisons seem to give a clear advantage to
> the sympy code. Think I'll go for that.
> 
> (as for brokenness of your patch for symbolic matrices, the problem
> isn't as much inexactness as undecidability of comparisons. Because of
> that, partial pivoting sticks with a fake zero pivot, which obviously
> fails).

Yes, that's what I found too.  My "solution" is to just stick with 
partial pivoting in the RR/RDF or CC/CDF cases.  Note that partial 
pivoting also fails in the padic case since the abs() function is not 
defined; hence the other kludge in the code.  That's why I decided it 
wasn't an exact/inexact thing, but rather a RDF/RR/CDF/CC versus 
anything else thing.

Thanks,

Jason


--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to