2

What is the domain of $f(x)=(\sqrt{x})^2$ ? Is it all real numbers, or are negative numbers still excluded, even after the square?

Edit: What I'm really wondering is whether $\lim\limits_{x \to 0} f(x)$ is defined. Sorry for the confusion.

  • Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. – Community Sep 04 '21 at 15:44
  • The limit is absolutely defined and it's zero. If you're wondering about complex values, it becomes more difficult to prove, but it's still (sorry but kinda obviously) zero. – Jared Sep 04 '21 at 16:00
  • @Jared I see. And I apologize for my ignorance. Does that mean that $\lim\limits_{x \to 0} \sqrt{x} = 0$ ? I couldn't find a straight answer online. – Jacob Lockard Sep 04 '21 at 16:01
  • You will find the correct answer (which is yes) here: https://math.stackexchange.com/questions/637280/limit-of-sqrt-x-as-x-approaches-0 – Hans Lundmark Sep 04 '21 at 16:03
  • @HansLundmark I had seen that post, but the answers seemed slightly uncertain. It seems that they must be correct, however. – Jacob Lockard Sep 04 '21 at 16:05
  • 1
    As they say, it depends a little on what definition you're using, but the answer is only no if you're using an unnecessarily restrictive definion (namely that the function must be defined in a punctured neighbourhood of the point). Here's a related question, which might also be of interest: https://math.stackexchange.com/questions/1914931/is-sqrt-x-continuous-at-0-because-it-is-not-defined-to-the-left-of-0 – Hans Lundmark Sep 04 '21 at 16:08
  • @JacobLockard So first off...if we're only concerned about real numbers, then you only need to show a right-handed limit (to zero) because there is no "left approach" of $\sqrt{x}$ in the reals. The complication comes when trying to prove this rigorously and you have to inspect all possible paths to $(0, 0)$ (where $x=$ real part and $y=$ imaginary part). – Jared Sep 04 '21 at 16:16

5 Answers5

6

If $A$ is a set and $f\colon A\longrightarrow\Bbb R$ is a function, then the domain of $f^2$ is equal to the domain of $f$, which is $A$. So, in particular, the domain of $\sqrt x^2$ is $[0,\infty)$, in spite of the fact that you always have $\sqrt x^2=x$.

Concerning $\lim_{x\to0}\sqrt x^2$, since $\sqrt x^2$ is defined only on $[0,\infty)$, it is equal to $\lim_{x\to0^+}x(=0)$ .

0

The maximal domain will be non negative real numbers if you consider $f(x) = g(x) \cdot g(x)$ where $g(x) = \sqrt x$ since the domain of $g$ is $[0,\infty)$.

But if you consider the function $h(x) = x$ then the domain of $h$ can be $\mathbb{R}$.

Infinity_hunter
  • 5,369
  • 1
  • 10
  • 30
0

Negative values for $x$ should be included for the function

$$h(x)=\sqrt{x^2}=|x|$$

but as noticed in this case $f(x)=\left(\sqrt {x}\right)^2$ stands for $f(x)=\sqrt x\sqrt x$ therefore negative values for $x$ are not allowed.

More precisely the given function is the composition $f=g\circ h$ of $h(x)=\sqrt x$ with $g(x)=x^2$ and by definition we first apply $h$ and then $g$.


edit

Therefore $\lim\limits_{x \to 0} f(x)$ is defined only as right side limit that is

$$\lim\limits_{x \to 0^+} f(x)$$

Usually we simply write

$$\lim\limits_{x \to 0} \sqrt x=0$$

implicitly assuming $x\ge 0$ that is $x\to 0^+$.

user
  • 154,566
  • So what about those who say otherwise, i.e. in https://math.stackexchange.com/questions/637280/limit-of-sqrt-x-as-x-approaches-0? Or am I misunderstanding something? – Jacob Lockard Sep 04 '21 at 16:08
  • It is mainly a fact of definition, when we write $\lim_{x\to 0} \sqrt x$ we mean $\lim_{x\to 0^+} \sqrt x$. – user Sep 04 '21 at 16:11
0

You can't "determine" a domain. The domain is given. The domain of $\sqrt{x}$ is anything you want it to be. What do you want? You want only real values (as the image)? Ok, then you must decide the domain is $x \in \mathbb{R}: x \geq 0$.

I will add, since I guess this is closer to your question: Is $\left(\sqrt{-1}\right)^2 = -1$? And does the fact that (it at least appears) we can have $x = -1$ and still get a real ($\mathbb{R}$) result. So does this "expand" the domain?

The answer is absolutely not. $\sqrt{-1} = i$ (and indeed $i^2 = -1$), so if you're afraid of complex numbers then you're still not allowed to use the negative square root.

Jared
  • 6,227
  • It is a quite common exercise type at a certain educational level to show an algebraic expression and ask the student to figure out for which values of the variable(s) the expression makes sense. Quibbling about the terminology used about this task ("find the domain of ...") is not really helpful for an OP who struggles to perform it in the first place. – Troposphere Sep 04 '21 at 16:04
  • @Troposphere I used to teach math and you're absolutely correct that this is a common exercise and one that gave me heartburn as a math teacher. I understand your critique (given expectations of a student), but I feel like I qualified my statement well enough: because what they don't tell you in high school is, what is the domain given these conditions on what we expect--which I feel as though I explained in my answer. – Jared Sep 04 '21 at 16:11
0

Fundamentally, what you are asking is: is $\lim\limits_{x\rightarrow 0}f^{-1}(f(x)) = 0$ because when you square the square root (you're "undoing" the square root). I find this to be a somewhat silly question because it comes down to domain (obviously $f^{-1}(f(x)) = x$ for all "normal" values of $x$). So then we have to wonder about "non-normal" values (e.g. $x = -0.0000001$ when thinking about $\sqrt{x}$).

This now becomes a vastly more complicated problem because an inverse is a strange thing: if a function is one-to-one (if I "know" $x$, then I "know" $y$--not the case when $f(x) = x^2$ because knowing that $f(x) = 1$ doesn't tell me the $x$ value--could be $x = 1$ or could be $x = -1$).

In this case, it's easy: $f^{-1}(f(x)) = x$ and we're done!

But what if this isn't the case--what if, like $\arcsin$ the inverse can give multiple solutions (I mean, the square root gives multiple solutions...so....)?

At that point, we look at the square root as "the problem" (it's really the inverse), so we don't actually have $f^{-1}(f(x))$ what we have is $f\left(f^{-1}(x)\right)$. Which means if $f(x)$ is well defined it doesn't matter what ridiculous value $f^{-1}(x)$ returns, $f\left(f^{-1}(x)\right) = x$, regardless (it's literally the definition of the inverse).

Edit

The square root is a "bad" function (because it's not a function at all)--it's an inverse function of the squared function ($f(x) = x^2$). You need to understand this. $f(x) = x^2$ isn't one-to-one which means the inverse isn't a function. A function must have: "one input" and "one output". The problem with the inverse of $x^2$ (the square root) is that it has two values--thus not a function.

Note

imo, the more "interesting" question is what is $\sqrt{x^2}$? And the answer is: $\sqrt{}$ is an inverse function and isn't well-defined (i.e. it's not a function). So we certainly can not say that $\sqrt{x^2} = x$ whereas we absolutely can say $\left(\sqrt{x}\right)^2 = x$.

Jared
  • 6,227