That's what I thought, but upon closer examination, it appears to only assign
the unary minus a higher priority on the first term, so y-exp(-x^2-z^2)
renders differently from y-exp(-z^2-x^2).  Good point about the sqr()
function, I hadn't realized that it was optimized.

> Without looking at the code, it looks like it's using too high a priority
> for unary minus, so it's treating "-x^2" as "(-x)^2".  BTW, I suspect you'd
> get better performance with the sqr() function.

Post a reply to this message