14

I'm trying to find a way to prove this:

EDIT: without using LHopital theorem. $$\lim_{x\rightarrow 1}\frac{x^{1/m}-1}{x^{1/n}-1}=\frac{n}{m}.$$ Honestly, I didn't come with any good idea.

We know that $\lim_{x\rightarrow 1}x^{1/m}$ is $1$.

I'd love your help with this.

Thank you.

  • I suggest playing around with various inequalities and using the squeeze theorem. For example, the inequality $(1+\epsilon)^a\leq 1+a\epsilon$ when $a,\epsilon >0$, helps quite a bit. – Cheerful Parsnip Apr 30 '11 at 21:19
  • 1
    Like Jim said, rather than the accepted answer, a more analytical way of thinking this would be using Taylor expansion: writing $x^{1/m} = (1 + \Delta x)^{1/m} = 1 + \frac{1}{m} \Delta x + o(\Delta x^2)$, the higher order terms will disappear when letting $\Delta x \rightarrow 0$ in that quotient you have. – Shuhao Cao May 03 '11 at 17:00
  • @ArturoMagidin I have changed the formatting of the title so as to make it take up less vertical space -- this is a policy to ensure that the scarce space on the main page is distributed evenly over the questions. See here for more information. Please take this into consideration for future questions. Thanks in advance. – GNUSupporter 8964民主女神 地下教會 Mar 19 '18 at 20:00
  • 1
    @GNUSupporter: I made that edit almost seven years ago. And if you take a look at my profile, perhaps you will see that your bit of helpful advice is... misguided in its intended target. I mean, really.... – Arturo Magidin Mar 19 '18 at 20:51
  • @ArturoMagidin You introduced \displaystyle to the title of this post, so I tagged you without looking at your profile. I do the same for every edit of this kind. You may take a look at my edit history to know more. – GNUSupporter 8964民主女神 地下教會 Mar 19 '18 at 20:57
  • @GNUSupporter: And given that you were addressing a 7 year old edit, perhaps you could have taken the few seconds to see if your scolding was perhaps a bit past the statute of limitations. By the way: taking a seven-year-old post and bumping it to the front page with an edit that addresses nothing but style is likewise rather frowned upon. Its most recent prior edit was a year ago. Perhaps you might also consider that next time you decide to do edits without doing much follow-up or thinking about them. – Arturo Magidin Mar 19 '18 at 22:38

5 Answers5

17

HINT $\ $ If you change variables $\rm\ z = x^{1/n} $ then the limit reduces to a very simple first derivative calculation. See also some of my prior posts for further examples of limits that may be calculated simply as first derivatives.

Bill Dubuque
  • 272,048
6

You can rewrite the limit as $$\lim_{x \rightarrow 1} {{x^{1\over m} - 1 \over x - 1} \over {x^{1 \over n} - 1 \over x- 1}}$$ By the quotient rule for limits this is exactly $${\lim_{x \rightarrow 1} {x^{1 \over m} - 1 \over x - 1} \over \lim_{x \rightarrow 1} {x^{1 \over n} - 1 \over x - 1}}$$ But notice that for any $\alpha$, ${\displaystyle \lim_{x \rightarrow 1} {x^{\alpha} - 1 \over x - 1}}$ is just the limit of difference quotients giving the definition of the derivative of the function $x^{\alpha}$ when evaluated at $x = 1$. So the limit is $\alpha$. So the limit in this question will be ${\displaystyle {{1 \over m} \over {1 \over n}} = {n \over m}}$.

Zarrax
  • 44,950
1

One thing about limits is that, if they exist, the "speed" at which you approach them doesn't matter. That is to say, $\lim_{x\rightarrow 1}\frac{x^{1/m}-1}{x^{1/n}-1} = \lim_{x^{1/n}\rightarrow 1}\frac{x^{1/m}-1}{x^{1/n}-1} = \lim_{y\rightarrow 1}\frac{y^{n/m}-1}{y-1}$. If you then apply L'Hopital's rule, you should get your answer.

Alex Becker
  • 60,569
  • 2
    If you're going to apply L'Hopital, why not do it immediately? But instead of invoking L'Hopital in your last step, one can observe that your transformed limit is a derivative at 1, see Bill Dubuque's answer. – wildildildlife Apr 30 '11 at 21:08
0

Are you aware of L'Hôpital's rule? It is useful in evaluating the limits of fractions such as yours.

Emre
  • 2,783
  • Yeah, I can't use it in this stage. I need to teach this someone who can't use LHopital right now. –  Apr 30 '11 at 20:51
-1

$$\frac{x^{1/m}-1}{x^{1/n}-1} = \frac{e^{\log(x^{1/m})}-1}{e^{\log(x^{1/n})}-1} = \frac{e^{\frac1{m}\log(x)}-1}{e^{\frac1{n}\log(x)}-1} = \frac{e^{\frac1{m}\log(x)}-1}{\log(x)} \frac{\log(x)}{e^{\frac1{n}\log(x)}-1}$$

$$\lim_{x \rightarrow 1} \frac{x^{1/m}-1}{x^{1/n}-1} = \lim_{\log(x) \rightarrow 0} \frac{e^{\frac1{m}\log(x)}-1}{\log(x)} \frac{\log(x)}{e^{\frac1{n}\log(x)}-1} = \lim_{y \rightarrow 0} \frac{e^{\frac{y}{m}}-1}{y} \frac{y}{e^{\frac{y}{n}}-1} = \frac1{m} \frac1{\frac1{n}} = \frac{n}{m}$$

  • isn't this a really indirect way of saying "use $lim_{\epsilon\rightarrow 0} (1+\epsilon)^{1/m} = 1+(\epsilon/m)+O(\epsilon^2)$ –  May 01 '11 at 03:41
  • @samanwita: I have used the fact $\lim_{y \rightarrow 0} \frac{e^y-1}{y} = 1$ which is probably one of the first limits people are taught. –  May 01 '11 at 04:16
  • Did you mean that $x$ tends to 1 in the second row? –  May 03 '11 at 09:45
  • @Nir: Yes. Thanks. –  May 03 '11 at 16:36