On Jan 27, 2008 11:31 PM, Robert Dodier <robert.dodier at gmail.com> wrote:
> Stavros, I'm adding these examples to the testsuite, if that's OK with
> you.
>
Fine with me.
> Shouldn't rat(m)^^-1 also be a CRE matrix?
>
Yes. I noticed the bug but hadn't gotten round to documenting it, sorry.
> In contrast, invert(rm) does yield a CRE matrix; .... Maybe foo^^-1 should
> simply punt to invert(foo) ??
If I remember correctly, ^^-1 and invert use different methods, which are
appropriate in different cases. I'd rather have the methods be chosen via
YAAGV (yet another global variable) than via different operators.
> it seems like operations on CRE matrices should yield CRE's.
>
Agreed.
-s