rref returns wrong results for sparse matrices

[David Bateman]

Jaroslav Hajek wrote:
> I've tracked this problem down to the statement
> 	A ([pivot, r], c:cols) = A ([r, pivot], c:cols);
>
> which apparently confuses Octave when pivot == r. The attached
> changeset is a quickfix for this; however, according to what John
> observed there may be bugs buried more deeply in liboctave.
>
>   
I believe the attached is a fix for this issue that addresses it within
the sparse assignment operators.

Regards
David
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: patch7622
Url: https://www.cae.wisc.edu/pipermail/bug-octave/attachments/20080304/763e16b1/attachment-0001.ksh