Barton Willis wrote:
> What I think is silly is the fact that Maxima does an asksign for
> atan2(x,x), but
> not for atan2(c * x, x), where c is a real number (say c = sqrt(3)).
> Surely Maxima
> should either do an asksign for both atan2(x,x) and atan2(c * x, x), or
> not do an
> asksign for both atan2(x,x) and atan2(c * x, x).
Don't know the intent but simpatan2 seems to have a lot of special
cases. Some of them are for atan2(x,x), atan2(-x,x) and the
"super-special-case" atan2(1,sqrt(3)).
Maybe atan2 should just try to see if y/x is a number. If so, compute
atan(y/x), but then ask for the signs so the right quadrant can be used.
Then atan(x,x) and atan2(c*x,x) will both ask. This also has the nice
effect of removing the super-special case of atan2(1,sqrt(3)).
Ray