24 Feb
2016
24 Feb
'16
1:48 p.m.
On Feb 24, 2016, at 3:36 PM, Dan Asimov <asimov@msri.org> wrote:
Never mind, I doubt the posistive triangular matrices is closed under inverses.
But: Certainly the group of positive diagonal matrices would work. Anything else?
Yes. Here’s a hint: Instead of thinking about the group of matrices, think about the corresponding group of transformations of the positive orthant.
—Dan