2

Given the function $$ f(x,y)=\begin{cases}\big|1+xy^2\big|^\dfrac{1}{x^2+y^2} & \quad\hfill (x,y)\neq(0,0)\\\\ 1 &\quad\hfill (x,y)=(0,0) \end{cases} $$

investigate whether the function is continuous at $(0,0)$.

Usually, I claim $p\in\mathbb{R}$ such that $y=px$ , $x\rightarrow 0$, placing them in $\displaystyle \lim_{x\rightarrow0}f(x,y)$ and seeing how that works out. If $\displaystyle \lim_{x\rightarrow0}f(x,y)=f(0,0) $ then (according to how I was taught) the function is continuous at $(0,0)$.

In this specific exercise, I can't seem to solve it using methods I know, i.e the one explained above, or just choosing $y$ to be any variation of $x$ (e.g. $y=\sqrt x $).

So now I'm at a standstill in my thoughts.

Any hints/tips would be really helpful!

Thanks!

311411
  • 3,576
Aishgadol
  • 399

3 Answers3

2

The trick you have is not the definition of continuity at $(0,0)$. By setting $y=px$ with various values of $p$ and considering $\lim_{x\to 0}f(x,px)$, if any these limits do not agree or do not exist, you can tell that $f$ is not continuous at $(0,0)$. But even if they all equal to $f(0,0)$, you can not conclude that $f$ is continuous at the origin. See these two related questions:

Two variable function that's continuous on all linear paths, but nevertheless discontinuous

Example of a function that is not continuous at $(0, 0)$ but continuous when restricted to any curve approaching the origin


When $(x,y)\ne (0,0)$, rewrite the function as $$ f(x,y)=\exp\left(\frac{\ln|1+xy^2|}{x^2+y^2}\right) $$

Now by continuity of the exponential function, you want to investigate whether $$ \lim_{(x,y)\to(0,0)}\frac{\ln|1+xy^2|}{x^2+y^2}=0\;. $$ By Taylor's expansion of the natural logarithm, near $(0,0)$, you can write $$ \ln|1+xy^2|=xy^2+g(x,y) $$ for some function $g$ where $|g(x,y)|\le C|(xy^2)^2|$. But $\lim_{(x,y)\to(0,0)}\frac{g(x,y)}{x^2+y^2}=0$, so it suffices to look at $$ \lim_{(x,y)\to(0,0)}\frac{xy^2}{x^2+y^2} =\lim_{(x,y)\to(0,0)}x\cdot \frac{1}{(x/y)^2+1}\;. $$ But this is zero because $$ |\frac{1}{(x/y)^2+1}|\le 1\;. $$

  • I've never come across a way to rewrite functions in such manner, therefor this solution raises more misunderstandings than actual things I can understand, given the knowledge base I have (according to what I was taught), I've managed to solve this problem, I'll share the solution as a comment for further review – Aishgadol May 03 '22 at 15:17
  • I did not mean to offend you in any type of way, that's not why I'm on this platform. Nonetheless, thank you for dedicating the time & thought to writing this solution, even though I did not benefit from it, I'm sure someone who understands the subject in a deeper sense can make use of it. – Aishgadol May 03 '22 at 15:31
  • rewriting the function as $exp( \cdot \cdot \cdot)$ is the first part of me not understanding. The rest is just more non-understandable information for me. Regarding the solution I came up with- care to elaborate about why is it incorrect? – Aishgadol May 03 '22 at 15:36
  • @Aishgadol: do you know that $a^b=e^{\ln a^b}=e^{b\ln a}$? $\exp(x)$ is simply a different way to write the exponential function $e^x$.// For the second question, do you know the definition of $f$ being continuous at $(0,0)$? Do you have a textbook? –  May 03 '22 at 15:37
  • I do have a textbook, however it states that $f(x,y)$ is continuous at $a\in\mathbb{R^2}$ if $a$ is an accumulation point and $\lim_{(x,y)\rightarrow a}f(x,y)=f(a)$ – Aishgadol May 03 '22 at 15:43
2

Your approach is not correct, using paths is only allowed to prove that a limit does not exist or in some cases to improve our intuition as to where the function approaches around a given point, nothing more.

Now, by definition

  • $f(0,0)=1$

Also, setting the change of variables to polar coordinates $$\begin{cases}x=r\cos \theta,\\y=r\sin \theta\end{cases}$$ with $r\in \mathbb{R}^{+*}$ and $\theta\in [0,2\pi[$, we have

\begin{align*} \lim_{(x,y)\to (0,0)}\left|1+xy^{2}\right|^\frac{1}{x^{2}+y^{2}}&=\lim_{r\to 0}|1+r^{3}\cos\theta\sin \theta|^{\frac{1}{r^{2}}},\\ &=\lim_{r\to 0}e^{\log |1+r^{3}\cos\theta\sin \theta|^{\frac{1}{r^{2}}}},\\ &=e^{\displaystyle \lim_{r\to 0}\frac{\log| 1+r^{3}\cos \theta\sin \theta|\underset{r\to 0}{\longrightarrow 0}}{r^{2}\underset{r\to 0}{\longrightarrow 0}}},\quad \text{(L'Hôpital's rule)}\\ &=e^{0},\\ &=1 \end{align*} Therefore $\displaystyle \lim_{(x,y)\to (0,0)}f(x,y)=f(0,0)$ then $f$ is continuous function in $(0,0)$.

A. P.
  • 6,048
  • After further inspection I managed to understand what you did there, do you have any general tips on cases when proving continuity? (for example, how did you came to the conclusion that $x=rcos\theta , y=rcos\theta$ were the best change of variables?) I'm trying to wrap my head around the methods of proving continuity for a two variable function – Aishgadol May 05 '22 at 15:13
  • Well, in general, there is no mystery in the presented solution: I make a change of variable to polar coordinates as part of the trial-error and motivated by the appearance of $\frac{1}{x^{2}+y^{2}}$ and finally I adjust everything to see the possibility of applying L'Hôpital's rule, which in this case it's possible. Finally I just use the operational definition of continuity. If you agree with the solution presented, you can click on the accept answer button, if you have any additional questions I will be on the lookout. – A. P. May 05 '22 at 17:04
-2

I think I managed to prove that the function is indeed continuous at $(0,0)$.

Choosing $y=\sqrt x$ , $x\rightarrow0$ I was able to prove the following: $$ \lim_{(x,y)\rightarrow(0,0)}|1+xy^2|^\dfrac{1}{x^2+y^2}=\lim_{x\rightarrow0}(1+x^2)^\dfrac{1}{x^2+x}=\lim_{x\rightarrow0}((1+x^2)^\dfrac{1}{x^2})^\dfrac{x^2}{x^2+x}=e^0=1=f(0,0) $$

which proves $f$ is continuous in $(0,0)$

I'm not quite sure this is enough, though all similar exercises I've seen in this subject are pretty much the same like this.

Aishgadol
  • 399
  • 2
    This is incorrect. It does not prove that the function is continuous at $(0,0)$. –  May 03 '22 at 15:27
  • "I'm not quite sure this is enough". Then you should probably not put it as an answer. You could add it into your question post instead. –  May 03 '22 at 15:30
  • " though all similar exercises I've seen in this subject are pretty much the same like this." It is very likely that you have serious misconceptions about those "exercises". It is very common to show a function being not continuous at certain point by taking limits along different paths. But one cannot conclude continuity from such approaches. –  May 03 '22 at 21:17