3
$\begingroup$

I can't solve this problem and I would be glad to get some help: Is the function $f(x,y) = x\ln(x^2 + 3y^2)$ continuous at $(0,0)$ $f(0,0)=0$?

$\endgroup$
  • 5
    $\begingroup$ I think you have to provide $f(0,0)$ in order to get an answer to this question. $\endgroup$ – Jan Jul 13 '19 at 20:49
  • $\begingroup$ yes, thank you I added it to the post. $\endgroup$ – Sapir Shuker Jul 14 '19 at 13:15
4
$\begingroup$

You didn't say what the value of the function is at $(0,0)$, but if $f(0,0) = 0$ then the function is continuous at $(0,0)$. Here are some hints on why this is the case.

Useful fact $1)$: $$|x\ln(x^2 + 3y^2)| \leq |(x^2 + 3y^2)^{1 \over 2}\ln(x^2 + 3y^2)|$$ Useful fact $2)$: $$\lim_{r \rightarrow 0^+}r^{1 \over 2} \ln r = 0$$ Useful fact $3)$: The squeeze rule.

Now try to put these all together, or use variations of the above statements.

$\endgroup$
0
$\begingroup$

Hint:

Passing to polar coordinates gives $$|f(r,\phi)| = |r\cos(\phi)\cdot \ln(r^2(1+2\cos^2\phi))| \le r|\ln (r^2)| \xrightarrow{r\to 0} 0$$

since $r\mapsto\left|\ln r\right|$ is decreasing for small $r$ and $1+2\cos^2\phi \ge 1$.

Therefore $\lim_{(x,y) \to (0,0)} f(x,y) = 0$.

$\endgroup$
  • $\begingroup$ @TedShifrin It shows how to find the right family of paths one should consider. By passing to polar coordinates it is clear that the first term goes to $0$ when $r \to 0$, and that by picking $r =\frac{c}{\ln(\sin(2\phi))}$ the second term goes to different values. $\endgroup$ – mechanodroid Jul 13 '19 at 22:11
  • 1
    $\begingroup$ I disagree. It's just as natural to think about choosing $y=f(x)$ so that $x\ln f(x)$ has an arbitrary limit. I might agree that you would win if we had $\ln(x^2+y^2)$ ... :) $\endgroup$ – Ted Shifrin Jul 13 '19 at 22:15
  • $\begingroup$ @TedShifrin Well, answers to questions like these which involve polar coordinates are in general not to be taken very seriously anyway, as there are situations like these: math.stackexchange.com/q/753381/144766 You are right, of course, this answer is mostly just for fun. $\endgroup$ – mechanodroid Jul 13 '19 at 22:16
  • 1
    $\begingroup$ Well, because @herb steinberg convinced me that you and I are both wrong, I looked more carefully at your solution. First, how did you get $1+2\cos^2\phi = \sin 2\phi$? Next, as $r\to 0$, $e^{c/r}$ is certainly not in the domain of $\arcsin$. $\endgroup$ – Ted Shifrin Jul 13 '19 at 22:57
  • $\begingroup$ @TedShifrin It was completely wrong, thanks. This hopefully shows that the limit of $f$ at $(0,0)$ is $0$, modulo the situation in the link. $\endgroup$ – mechanodroid Jul 13 '19 at 23:08
0
$\begingroup$

$x→0$ faster than the ln term →−∞, so the $f(x,y)$ is continuous at $(0,0)$ (proof follows). In fact $f(0,y)=0$ for any value of $y$. In case there is any uncertainty about $f(0,0)$ being undefined, just call it $=0$ at that point. .

Without loss of generality assume $x^2+3y^2\lt 1$

For $x\gt 0$, $xln(x^2)\lt xln(x^2+3y^2)\lt 0$.

For $x\lt 0$, $xln(x^2)\gt xln(x^2+3y^2)\gt 0$. In both cases $xlnx^2\to 0$ as $x\to0$, so that $f(x,y)\to 0$.

.

Therefore $f(x,y)$ is continuous at $(0,0)$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.