Re: invariance of negative signature of the metric?



On Mar 11, 11:32 am, "Ken S. Tucker" <dynam...@xxxxxxxxxxxx> wrote:

Can you find a reason to exclude

&x / &x' = sqrt(-1)

as a legit transformation coefficient?

I'm sure that it's possible to work perfectly well with
complex coordinates. But the notion of a metric "signature"
only makes sense if the coordinates are real.

--
Daryl McCullough
Ithaca, NY

.