-1
$\begingroup$

If limit exists, then what is its value? And if it does not exist then can we find where does this expression tends as $ n \to \infty$.

The expression : $\lim\limits_{n \to \infty } \sum\limits_{k=1}^n - (-1)^{k} \ln \left(\frac{1}{k}\right)$

$\endgroup$

2 Answers 2

0
$\begingroup$

For a sum $\sum_{k=1}^\infty a_k$ to converge, a necessary condition is that

$$\lim_{k \to \infty} a_k = 0,$$

which is not the case here. Therefore, your sum diverges.

$\endgroup$
1
  • $\begingroup$ We can write the above expression as $-\ln(1)+\ln(2)-\ln(3)+\ln(4)-\ln(5)...\infty$ and $\ln(1) = 0$ , thus it becomes : $\ln(2)-\ln(3)+\ln(4)-\ln(5)-\ln(6)...\infty$ which can be written as $\ln\left(\frac{2}{3}\right) + \ln\left(\frac{4}{5}\right) + \ln\left(\frac{6}{7}\right) + \ln\left(\frac{8}{9}\right) + \ln\left(\frac{10}{11}\right) + .... \ln\left(\frac{n-1}{n}\right)$; and as $n\to\infty$, $\lim\limits_{n\to\infty}\left(\frac{n-1}{n}\right) = 1$ thus $\lim\limits_{n \to \infty}ln\left(\frac{n-1}{n}\right)=0$ , that is $\lim\limits_{k\to\infty} a_k = 0$, so it should converge. $\endgroup$
    – Dost Arora
    Aug 13, 2015 at 12:40
-2
$\begingroup$

The sum can be simplified as

$\sum\limits_{k=1}^n - (-1)^{k} \ln \left(\frac{1}{k}\right) =\sum\limits_{k=1}^n (-1)^{k} \ln k $.

This sum diverges. However, let's see how it behaves for even and odd numbre of terms.

For an even number of terms,

$\begin{array}\\ S_{2n} &=\sum\limits_{k=1}^{2n} (-1)^{k} \ln k\\ &=\sum\limits_{k=1}^{n} (- \ln (2k-1)+\ln(2k))\\ &=\sum\limits_{k=1}^{n} (- \ln (\frac{2k-1}{2k})\\ &=-\sum\limits_{k=1}^{n} \ln (1-\frac{1}{2k})\\ &\approx-\sum\limits_{k=1}^{n} -(\frac{1}{2k}+\frac{1}{8k^2}+ ...)\\ &=\sum\limits_{k=1}^{n} \frac{1}{2k}+\sum\limits_{k=1}^{n}(\frac{1}{8k^2}+ ...)\\ &=\frac12 H_n+C \quad\text{for some constant } C \text{ and large } n\\ &\approx\frac12 (\ln(n)+\gamma)+C\\ &=\frac12 \ln(n)+C_1\\ \end{array} $

Note that $C_1 =C+\frac12 \gamma $ and $C = \sum_{k=1}^{\infty} \sum_{m=2}^{\infty}\frac1{m(2k)^m } = \sum_{m=2}^{\infty}\frac1{m2^m}\sum_{k=1}^{\infty} \frac1{ k^m } = \sum_{m=2}^{\infty}\frac1{m2^m}\zeta(m) $ which converges and is easily computed.

For an odd number of terms,

$\begin{array}\\ S_{2n+1} &= S_{2n}-\ln(2n+1)\\ &\approx \frac12 \ln(n)+C_1-\ln(2n+1)\\ &= \frac12 \ln(n)+C_1-(\ln 2 + \ln(n+\frac12))\\ &= \frac12 \ln(n)+C_1-(\ln 2 + \ln(n)+\ln(1+\frac1{2n}))\\ &= -\frac12 \ln(n)+C_1-\ln 2+\frac1{2n}+O(\frac1{n^2}))\\ &\approx -\frac12 \ln(n)+C_1-\ln 2\\ \end{array} $

If we average the even and odd sums, $\frac12(S_{2n}+S_{2n+1}) \approx C_1-\frac12\ln 2 $.

This can be considered as the limit of the sum and $C_1$ can be computed as described previously.

$\endgroup$
6
  • $\begingroup$ Thank You very much for your answer, but I can't understand how does it diverge. $\endgroup$
    – Dost Arora
    Aug 12, 2015 at 7:04
  • $\begingroup$ Since $\ln (1/k) =-\ln(k)$, the terms of the sum are unbounded. Perhaps you meant $\ln(1-1/k)$. If so, please ask this as another question instead of editing this one. If my guess is correct, look up the Wallis product. $\endgroup$ Aug 12, 2015 at 13:45
  • $\begingroup$ Why the downvote? I produced a nonstandard summation method which gave a value. $\endgroup$ Aug 12, 2015 at 19:39
  • $\begingroup$ I didn't do it. $\endgroup$
    – Dost Arora
    Aug 13, 2015 at 9:53
  • $\begingroup$ We can write the above expression as $-\ln(1)+\ln(2)-\ln(3)+\ln(4)-\ln(5)...\infty$ and $\ln(1) = 0$ , thus it becomes : $\ln(2)-\ln(3)+\ln(4)-\ln(5)-\ln(6)...\infty$ which can be written as $\ln\left(\frac{2}{3}\right) + \ln\left(\frac{4}{5}\right) + \ln\left(\frac{6}{7}\right) + \ln\left(\frac{8}{9}\right) + \ln\left(\frac{10}{11}\right) + .... \ln\left(\frac{n-1}{n}\right)$; and as $n\to\infty$, $\lim\limits_{n\to\infty}\left(\frac{n-1}{n}\right) = 1$ thus $\lim\limits_{n \to \infty}ln\left(\frac{n-1}{n}\right)=0$ , so it should converge. $\endgroup$
    – Dost Arora
    Aug 13, 2015 at 12:37

You must log in to answer this question.

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