steve_bank said:
It is pointless to try to argue theory when you do not understand actual math that is use.
Fine, I left the condition that (sqrt(x))^2=x implicit before. My bad. So, let D be the domain of sqrt, and let's say that for all x in D, (sqrt(x)^2)=x and sqrt(x)sqrt

=sqrt(xy). Let us also stipulate that sqrt(x) takes the usual values for all non-negative real x. Then,
-1=sqrt(-1))^2=sqrt((-1)^2)=sqrt(1)=1.
Thus,
no such function exists.
It does not exist using real numbers. Limited to reals there is no solution to the square root of a negative number.
sqr(-4 * 9) = sqr(-1) * sqr(4) * sqr(9) = [0 + i6]...i6 js on the imaginary not real axis.
[0 + i6]^2= [0 + i6] * [0 + i6] = i^2 * 6^2 = -1*36 = -4 * 9
In electrical theory and electronics in general square roots of negative numbers are routine. It requires complex numbers. Reals are a subset of complex numbers.
The solution is not on the real number line.
There are an infinite number of integer roots.
(x)^1/n n = 0,1,2,3....
(-1)^n has a real solution for n = 1,3,5.... and it is -1.