Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I am trying to find the limit of the following integral as $n$ grows to infinity.

$$\int_1^{ n+1 }\frac { \ln x }{ x^n } \, dx $$

However, when I am doing an integration by parts, the result looks ugly. It is not in a simple form from which, I think, one can find the limit in a simple way (indeterminate form, I mean).

So there should be a trick in there to find the limit.

Can someone help me on this one please?

Regards,

share|improve this question
3  
Why does the result of integration by parts look ugly? You get a pretty simple thing. –  Daniel Fischer Sep 4 '13 at 22:09
    
Daniel, just try using Wolfram Alpha, to see the look, it is not a simple thing... –  XCoder Sep 4 '13 at 22:14

5 Answers 5

up vote 3 down vote accepted

$$0<\int_1^{ n+1 }\frac { \ln x }{ x^n } \, dx<\int_1^{\infty}\frac { \ln x }{ x^n } \, dx < \int_1^{\infty }x^{1-n } \, dx=\frac{1}{n-2}$$ but $\lim_{n\to\infty}\frac{1}{n-2}=0$, so by the Squeeze Theorem your integral is $0$.

share|improve this answer

Let $u=\ln x$ and $dv=x^{-n}\,dx$. Then $du=\frac{1}{x}\,dx$ and we can take $v=\frac{1}{-n+1}x^{-n+1}$.

The second integration will be easy.

Remark: One can avoid some of the complications by noting the fact that for $x\ge 1$ we have $0\le \ln x\le x-1$.

share|improve this answer
    
That's what I did, André, but the final result is not that simple. I did not reproduce here, because I have trouble typing long equations using MathJax. But looking at the result from Wolfram Alpha, it is not so easy to derive a limit here. Result: (1 - (1 + n)^(1 - n) (1 + (-1 + n) Log[1 + n]))/(-1 + n)^2 –  XCoder Sep 4 '13 at 22:15
    
If you want slightly less ugly, you can use the fact that for $x\ge 1$, we have $0\le \ln x\le x-1$. –  André Nicolas Sep 4 '13 at 22:25
    
For me, this thing always tends to 0 as n grows. Am I correct in saying this ? –  XCoder Sep 4 '13 at 22:26
    
Yes, the limit is $0$. If one has experience, it is obvious. Pick a small $\epsilon$. The integral from $\epsilon$ up goes to $0$, because $\frac{1}{x^n}$ shrinks rapidly. The integral from $1$ to $1+\epsilon$ is also small, because on this interval $\ln x$ is close to $0$. –  André Nicolas Sep 4 '13 at 22:29
    
There are also fancy theorems one could quote, but I am guessing the course is an early one. –  André Nicolas Sep 4 '13 at 22:36

$$ \begin{align} & \int_1^{n+1} \underbrace{\Big(\ln x\Big)}_{u} \underbrace{\left(\frac{1}{x^n} \, dx\right)}_{dv} = \int u\,dv = uv - \int v\,du \\[12pt] & = \left.(\ln x)\left(\frac{x^{-n+1}}{-n+1}\right)\right|_1^{n+1} - \int_1^{n+1} \left(\frac{x^{-n+1}}{-n+1}\right) \frac{dx}{x} \\[12pt] & = \frac{\ln(n+1)}{-n+1}\cdot\frac{1}{(n+1)^{n-1}} + \frac{1}{n-1} \int_1^{n+1} x^{-n}\,dx \\[12pt] & = \frac{1}{n-1}\cdot\frac{1}{(n+1)^n} + \frac{1}{n-1}\cdot\left(\frac{1}{(n+1)^n}-1\right) \to 0 \text{ as }n\to\infty. \end{align} $$

share|improve this answer
    
I think you forgot about your $\ln(n+1)$ term while moving from line $3$ to line $4$? –  pre-kidney Sep 4 '13 at 23:38

I am seeing a whole lot of zero here. If you do the manual integration then evaluate the result term by term there are -n exponents on all of the terms in the numerators. If you expand your result in mathematica you get something like

$$-\frac{(n+1)^{1-n}}{(n-1)^2}+\frac{1}{(n-1)^2}-\frac{n (n+1)^{1-n} \log (n+1)}{(n-1)^2}+\frac{(n+1)^{1-n} \log (n+1)}{(n-1)^2}.$$

(I just used Expand[] on your result, did TeXForm[%] on that and pasted that result in here between dollar symbols.)

Anyway, as $n \rightarrow \infty$, the numerators (note the -n exponents) approach zero, the denominators diverge, and so the limit of the integral is zero.

share|improve this answer
1  
One can say either the sequence approaches zero or the limit is zero, but you shouldn't say the limit approaches zero. –  Michael Hardy Sep 4 '13 at 22:53
    
I will buy that. Thanks, and I must say your derivation is top notch writing. –  J. W. Perry Sep 4 '13 at 22:56
    
Thank you. ${{{{}}}}$ –  Michael Hardy Sep 4 '13 at 23:03
    
Thanks to all for your help. –  XCoder Sep 4 '13 at 23:38

\begin{align} \int_{1}^{n + 1}{\ln\left(x\right) \over x^{n}}\,{\rm d}x &= \lim_{m \to 0} {\partial \over \partial m}\int_{1}^{n + 1}{x^{m} \over x^{n}}\,{\rm d}x = \lim_{m \to 0} {\partial \over \partial m} \left\lbrack\left(n + 1\right)^{m - n + 1} - 1 \over m - n + 1\right\rbrack \\[3mm]&= \lim_{m \to 0} {\left(n + 1\right)^{m - n + 1}\ln\left(n + 1\right)\,\left(m - n + 1\right) - \left\lbrack\left(n + 1\right)^{m - n + 1} - 1\right\rbrack \over \left(m - n + 1\right)^{2}} \\[3mm]&= {\left(n + 1\right)^{1 - n}\,\ln\left(n + 1\right)\,\left(1 - n\right) - \left(n + 1\right)^{1 - n} + 1 \over \left(n - 1\right)^{2}} \\[5mm]--------&------------------------------- \end{align}

\begin{align} \lim_{n \to \infty}\int_{1}^{n + 1}{\ln\left(x\right) \over x^{n}}\,{\rm d}x &= \lim_{n \to \infty} {n^{-n}\,\ln\left(n\right)\,\left(-n\right) - n^{-n} + 1 \over n^{2}} \\[3mm]&= -\lim_{n \to \infty}{\ln\left(n\right) \over n^{n + 1}} + \lim_{n \to \infty}{1 \over n^{n + 2}} + \lim_{n \to \infty}{1 \over n^{2}} \\[5mm]& \end{align}

$$ \begin{array}{|c|}\hline \\ \color{#ff0000}{\large\quad% \lim_{n \to \infty}\int_{1}^{n + 1}{\ln\left(x\right) \over x^{n}}\,{\rm d}x = 0\quad} \\ \\ \hline \end{array} $$

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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