There is a rule that if a function is not one to one, then its inverse is not a function.
When graphed, a quadratic function is not one to one.
However, there is also a rule that the square root and radical sign with the default index of 2 only refer to the positive square root.
Is the square root function the inverse of a quadratic function?
If so, why are the quadratic and the square root functions special exceptions to this rule?
I have asked an accidentally duplicate question about why the square root symbol refers to only the positive one and it seems like it is only due to mathematical convention. Is it the same reason why?
Any help with explaining is greatly appreciated.


So you're right that there is a problem. The square root function is not an inverse to the function $f(x)=x^2$ on its domain. You can see that it doesn't satisfy the main thing that we want from an inverse: that $g(f(x))$ is equal to $x$. For example if we start with $-4$, and then put it first in $f(x)=x^2$ and then $g(x)=\sqrt{x}$, we get $\sqrt{(-4)^2} = \sqrt{16} = 4.$
The best that we can do here is take a branch of the function $f(x)=x^2$ where it does happen to be one-to-one. If we choose the set $[0,\infty)$ we then can take the inverse of that restricted function. That ends up being the function we know as $\sqrt{x}.$
This isn't that big of an issue so long as we are aware of what we are doing. The inverse trig functions such as $\arcsin(x)$ and $\arctan(x)$ all are inverses of a branch of the original function. For example if you take $\arctan(\tan(\pi))$ you will get $0$ instead of $\pi$.
There are some famous actual inverse functions in calculus though. $e^x$ and $\ln x$ are perhaps the most famous examples.