-2
$\begingroup$

$$\displaystyle\lim_{(x,y)\rightarrow (1,0)}\frac{x-1}{\sqrt{(x-1)^2+y^2}}$$

By direct substitution that's a ( $ \frac{0}{0}$ ) undefined

so can I approach it by polar equation or by different paths ?

Mine was the polar form considering eqaution of circle is :

$$a(x-x_1)^2+b(y-y_1)^2$$

$\endgroup$
3
  • $\begingroup$ Mayar: the editor made the assumption that the entire demominator is under the square root site. What you wrote suggests you could have also meant the denominator to be $\sqrt{(x-1)^2} + y^2$. Can you please either confirm the edit, or else the variation I've just mentioned. Your original post was ambiguous. $\endgroup$
    – amWhy
    Sep 25, 2018 at 20:50
  • $\begingroup$ It is as the editor did the whole denomirator under the root $\endgroup$
    – M12567
    Sep 25, 2018 at 20:57
  • $\begingroup$ HINT wlog we can have the limit $ \dfrac{x}{r}= \cos \theta$ after translation $\endgroup$
    – Narasimham
    Sep 26, 2018 at 4:28

3 Answers 3

3
$\begingroup$

Let $x=u+1, \, v=y$ with $(u,v)\to (0,0)$ then

$$\lim_{(x,y)\rightarrow (1,0)}\frac{x-1}{\sqrt{(x-1)^2+y^2}}=\lim_{(u,v)\rightarrow (0,0)}\frac{u}{\sqrt{u^2+v^2}}$$

and since by polar coordinates $\frac{u}{\sqrt{u^2+v^2}}=\cos \theta$ we can conclude that the limit doesn't exist, indeed

  • $u=0 \implies \frac{u}{\sqrt{u^2+v^2}}=0$

  • $u=v \implies \frac{u}{\sqrt{u^2+v^2}}=\frac{\sqrt 2}2$

$\endgroup$
1
$\begingroup$

Different paths are useful in this exercise, easy to choose and compute.

If $x=1$ you get $0.$

If $x>1$ and $y=0,$ the result is $1.$

This is enough to conclude that the limite does not exist.

$\endgroup$
3
  • $\begingroup$ For the second path maybe $x>1$ or $x< 1$ is more indicated and the limit is $\pm 1$. For that reason I’ve suggested a change of variables, it is not necessary but it is recommended in order to avoid such kind of mistakes. In general indeed it is always better to refer to limit at the origin or at $\infty$. We are not forced to fo that explicitly but it is better at lest check the result in that way. $\endgroup$
    – user
    Sep 26, 2018 at 2:56
  • $\begingroup$ Sorry maybe I was not clear, but you line "If $x>0$ and $y=0,$ the result is $1$" is not correct. $\endgroup$
    – user
    Sep 26, 2018 at 6:49
  • $\begingroup$ Right, I am editing now. $\endgroup$
    – user376343
    Sep 26, 2018 at 7:26
1
$\begingroup$

Put

$$x=1+r\cos(t)$$ and

$$y=r\sin(t)$$

the function becomes

$$\frac{r\cos(t)}{r}=\cos(t)$$

the limit depends on the angle $t$, so The limit doesn't exist.

$\endgroup$
6
  • $\begingroup$ If applied the different path method it brings a not existing limit , could you please show me another workout using the different paths method in proving the limit's existence $\endgroup$
    – M12567
    Sep 25, 2018 at 20:58
  • $\begingroup$ The different paths method serves to prove non-existence. To prove the existence, a different method is needed. $\endgroup$
    – user376343
    Sep 25, 2018 at 21:30
  • $\begingroup$ So how it doesnot exist by the polar coordinates ? after$$cos(t)$$ $\endgroup$
    – M12567
    Sep 25, 2018 at 21:32
  • $\begingroup$ $\cos(t)$ is not a constant, gets different values ... $\endgroup$
    – user376343
    Sep 25, 2018 at 21:35
  • 1
    $\begingroup$ oooh okay its not an existing limit bec yes because it doen't equal one constant value , yes I get it now , thank you $\endgroup$
    – M12567
    Sep 25, 2018 at 21:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .