3
$\begingroup$

I'm reading a proof and very rusty. I'm missing something simple I suspect.

Have a little mercy and give me a nudge in the right direction?

$2\epsilon + \sum_{k=2}^{\infty} 2\epsilon \left( \frac{1}{k-1} - \frac{1}{k} \right) = 4\epsilon$

I can use $\sum k^{-2}=\frac{\pi^2}{6}$ to show that it is less than $6\epsilon$ which also works for the proof this is from, but how do I show it is exactly $4\epsilon$ since that is how the book does it? In other words how do I show:

$ \sum_{k=2}^{\infty} \left( \frac{1}{k-1} - \frac{1}{k} \right) = 1$

I would prefer a hint.

$\endgroup$
  • 2
    $\begingroup$ Take the partial sums and see what's getting canceled ... $\endgroup$ – rtybase Feb 4 '17 at 16:34
  • 2
    $\begingroup$ So (if you're very rusty): $\frac{1}{1} - \frac{1}{2} + \frac{1}{2} - \frac{1}{3} + \frac{1}{3} - \frac{1}{4}... + \frac{1}{n-1} - \frac{1}{n}$ e.g. equals $1 - \frac{1}{n}.$ $\endgroup$ – Obriareos Feb 4 '17 at 16:38
  • 1
    $\begingroup$ what a ridicolous downvote $\endgroup$ – tired Feb 4 '17 at 16:41
  • $\begingroup$ @Obriareos Your comment is effectively the answer I posted. $\endgroup$ – Mark Viola Feb 4 '17 at 16:43
  • $\begingroup$ @ tired: Downvote? @ Dr. MV: that's true - I haven't see your answer before. $\endgroup$ – Obriareos Feb 4 '17 at 19:11
5
$\begingroup$

Note that

$$\begin{align} \sum_{k=2}^K\left(\frac{1}{k-1}-\frac1k\right)&=\left(1-\frac12\right)+\left(\frac12-\frac13\right)+\left(\frac13-\frac14\right)+\cdots +\left(\frac{1}{K-1}-\frac{1}{K}\right)\\\\ &=1-\frac1K \tag 1 \end{align}$$

More formally, note that if $S_K=\sum_{k=2}^K\left(\frac{1}{k-1}-\frac1k\right)$, then $S_{K+1}=S_K+\frac{1}{K}-\frac{1}{K+1}$. So, we can use induction to prove the result given by $(1)$.

$\endgroup$
  • $\begingroup$ Hehe ...too many very similar to a basic question. I'm deleting mine...+1 $\endgroup$ – DonAntonio Feb 4 '17 at 16:36
  • $\begingroup$ @DonAntonio Yes indeed. $\endgroup$ – Mark Viola Feb 4 '17 at 16:37
17
$\begingroup$

Hint:

enter image description here

hope you can take it from here.

As @CarstenS pointed out the picture shown here is somehow a more correct hint

$\endgroup$
  • 6
    $\begingroup$ Now that deserves a (+1) just for the creativity! $\endgroup$ – Mark Viola Feb 4 '17 at 16:38
  • 1
    $\begingroup$ @Oliver821 very good! $\endgroup$ – tired Feb 4 '17 at 16:38
  • 1
    $\begingroup$ @Dr.MV thank you very much! :) $\endgroup$ – tired Feb 4 '17 at 16:39
  • 1
    $\begingroup$ Yes the hint should be obvious by now. Hint: Split the sums and solve each telescoping series individually. $\endgroup$ – John Smith Feb 4 '17 at 16:39
  • 2
    $\begingroup$ And this put a big smile on my face. Nice picture by the way. $\endgroup$ – Mark Viola Feb 4 '17 at 16:43
1
$\begingroup$

It's a telescoping series.

First show:

$$\sum_{k=2}^N \frac{1}{k-1}-\frac{1}{k}=1-\frac{1}{N}$$

$\endgroup$
1
$\begingroup$

Fix $n\geq 2$. Then $$\sum_{k=2}^n\left(\frac{1}{k-1}-\frac{1}{k}\right)=\sum_{k=2}^n \frac{1}{k-1}-\sum_{k=2}^n\frac{1}{k}=\sum_{k=1}^{n-1}\frac{1}{k}-\sum_{k=1}^n\frac{1}{k}=1+\sum_{k=2}^{n-1}\frac{1}{k}-\sum_{k=2}^{n-1}\frac{1}{k}-\frac{1}{n}=1-\frac{1}{n}.$$ Let $n\to \infty $ and you'll get your result.

$\endgroup$

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.