What if we rearrange the terms of a known series?

Calculus Level 5

1 2 + 1 4 1 1 + 1 6 + 1 8 1 3 + 1 10 + 1 12 1 5 + \frac{1}{2}+\frac{1}{4}-\frac{1}{1}+\frac{1}{6}+\frac{1}{8}-\frac{1}{3}+\frac{1}{10}+\frac{1}{12}-\frac{1}{5}+\cdots

The series above is obtained by rearranging the terms of the alternating harmonic series. Determine the approximate value of the series rounded to three decimal places if it converges; otherwise, enter 0.


The answer is -0.347.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Otto Bretscher
May 3, 2016

Combining each subtracted term with the term two places before, we can contract the series to 1 2 + 1 4 1 6 + 1 8 . . . = ln ( 2 ) 2 0.347 -\frac{1}{2}+\frac{1}{4}-\frac{1}{6}+\frac{1}{8}-...=-\frac{\ln(2)}{2}\approx \boxed{-0.347} , negative half of the alternating harmonic series.

Any reason why the following is wrong? 1 2 + 1 4 1 + 1 6 + 1 8 1 3 + 1 10 + 1 12 1 5 + . . . = 1 2 + 1 4 + 1 6 + . . . ( 1 + 1 3 + 1 5 + 1 7 + . . . ) \frac{1}{2}+\frac{1}{4}-1+\frac{1}{6}+\frac{1}{8}-\frac{1}{3}+\frac{1}{10}+\frac{1}{12}-\frac{1}{5}+... = \frac{1}{2}+\frac{1}{4}+\frac{1}{6}+...-\left(1+\frac{1}{3}+\frac{1}{5}+\frac{1}{7}+...\right) = 1 2 ( 1 1 + 1 2 + 1 3 + . . . ) ( 1 + 1 3 + 1 5 + 1 7 + . . . ) =\frac{1}{2}\left(\frac{1}{1}+\frac{1}{2}+\frac{1}{3}+...\right)-\left(1+\frac{1}{3}+\frac{1}{5}+\frac{1}{7}+...\right) = 1 2 ( 1 2 + 1 4 + 1 6 + . . . ) + 1 2 ( 1 + 1 3 + 1 5 + 1 7 + . . . ) ( 1 + 1 3 + 1 5 + 1 7 + . . . ) =\frac{1}{2}\left(\frac{1}{2}+\frac{1}{4}+\frac{1}{6}+...\right)+\frac{1}{2}\left(1+\frac{1}{3}+\frac{1}{5}+\frac{1}{7}+...\right)-\left(1+\frac{1}{3}+\frac{1}{5}+\frac{1}{7}+...\right) = 1 2 ( 1 2 + 1 4 + 1 6 + . . . ) 1 2 ( 1 + 1 3 + 1 5 + 1 7 + . . . ) = half of the orginal expression =\frac{1}{2}\left(\frac{1}{2}+\frac{1}{4}+\frac{1}{6}+...\right)-\frac{1}{2}\left(1+\frac{1}{3}+\frac{1}{5}+\frac{1}{7}+...\right)=\text{half of the orginal expression} Then, using this logic, I got an answer of 0. Any reason why this is wrong?

A Former Brilliant Member - 5 years, 1 month ago

Log in to reply

I think you should do the same thing, but working with partial sums. Notice that the alternating harmonic series is not absolutely convergent (it is conditionally convergent), so when you rearrange the terms of the series, you might change its sum. You can consider a partial sum and rearrange the terms in a convenient way, as I did in my solution. You can take a look of the wiki about absolute convergence.

Arturo Presa - 5 years, 1 month ago

It is a common rookie mistake / misconception made by those who are not familiar with calculus / analysis. As pointed out by Arturo, in order for us to justify rearranging terms, we have to ensure that sequence is absolutely convergent .

If the sequence is not absolutely convergent, then we can possibly rearrange the terms to get another limit, as you did above.

Calvin Lin Staff - 5 years, 1 month ago

Log in to reply

Thank you, I understand now.

A Former Brilliant Member - 5 years, 1 month ago

Very nice solution, Otto! I gave you my vote!

Arturo Presa - 5 years, 1 month ago

Ya.. ( + 1 ) (+1) ...same way ...... 1 / 2 -1/2 times expansion series of ln ( 1 + x ) \ln(1+x) at x = 1 x=1 .

PS:- I was going to write a one line solution for this problem but found you already here.. :-)

Rishabh Jain - 5 years, 1 month ago

Log in to reply

My solution is a lengthy two lines, Comrade... go ahead and write your one-liner! ;)

Otto Bretscher - 5 years, 1 month ago

Log in to reply

No..... My solution is essentially same as yours.... I would obviously skip some texts.... :-)

Rishabh Jain - 5 years, 1 month ago

Log in to reply

@Rishabh Jain Instead, go ahead and write a one line solution to this one ;)

Otto Bretscher - 5 years, 1 month ago

Log in to reply

@Otto Bretscher I would try my very best... !!

Rishabh Jain - 5 years, 1 month ago

The series is convergent, Hence, the terms can be rearranged into 1 + 1 2 1 3 + 1 4 + = ln ( 2 ) -1+\frac{1}{2}-\frac{1}{3}+\frac{1}{4}+\cdots=-\ln(2)

How come the same convergent series has two solutions. Particularly, when the brute force summation does converge to ln ( 2 ) -\ln(2) ??

Janardhanan Sivaramakrishnan - 5 years, 1 month ago

Log in to reply

Your premise, "the series is convergent, hence, the terms can be rearranged", is flawed, as Mr. Presa explains in his comment. You need absolute convergence.

Otto Bretscher - 5 years, 1 month ago
Arturo Presa
May 3, 2016

Let us use the notation H n = 1 + 1 2 + 1 3 + . . . + 1 n . H_n=1+\frac{1}{2}+\frac{1}{3}+...+\frac{1}{n}. The number H n H_n is called the n t h n^{th} - harmonic number. If we add the first 3 n 3n first terms of the given series, we obtain the following expression S 3 n = k = 1 2 n 1 2 k k = 1 n 1 2 k 1 = k = 1 2 n 1 2 k k = 1 n 1 2 k 1 k = 1 n 1 2 k + k = 1 n 1 2 k S_{3n}=\sum_{k=1}^{2n} \frac{1}{2k}-\sum_{k=1}^{n}\frac{1}{2k-1}=\sum_{k=1}^{2n} \frac{1}{2k}-\sum_{k=1}^{n}\frac{1}{2k-1}-\sum_{k=1}^{n}\frac{1}{2k}+\sum_{k=1}^{n}\frac{1}{2k} = k = 1 2 n 1 2 k k = 1 2 n 1 k + k = 1 n 1 2 k . =\sum_{k=1}^{2n} \frac{1}{2k}-\sum_{k=1}^{2n}\frac{1}{k}+\sum_{k=1}^{n}\frac{1}{2k}. It is easy to see that S 3 n = 1 2 H 2 n + 1 2 H n . S_{3n}=-\frac{1}{2}H_{2n}+\frac{1}{2}H_n. Therefore, we obtain that S 3 n = 1 2 ( H 2 n ln 2 n ) + 1 2 ( H n ln n ) 1 2 ln 2 n + 1 2 ln n S_{3n}=-\frac{1}{2}(H_{2n}-\ln{2n})+\frac{1}{2}(H_n - \ln n)- \frac{1}{2}\ln{2n} +\frac{1}{2}\ln n = 1 2 ( H 2 n ln 2 n ) + 1 2 ( H n ln n ) + 1 2 ln 1 2 =-\frac{1}{2}(H_{2n}-\ln{2n})+\frac{1}{2}(H_n - \ln n)+\frac{1}{2}\ln{\frac{1}{2}} Now using that lim n ( H n ln n ) = γ , \lim_{n\rightarrow \infty}(H_n-\ln n)=\gamma, where γ \gamma is the Euler–Mascheroni constant, we obtain that lim n S 3 n = 1 2 ln 1 2 0.347. \lim_{n\rightarrow \infty} S_{3n}=\frac{1}{2}\ln{\frac{1}{2}}\approx -0.347.

There is one detail that is missing in my solution. Can somebody tell?

Arturo Presa - 5 years, 1 month ago

Log in to reply

You have to make sure that the sum of S 3 n + 1 , S 3 n + 2 S_{3n+1},S_{3n+2} approaches to that of S 3 n S_{3n} , which we can get from the squeeze theorem.

Noam Pirani - 5 years, 1 month ago

Log in to reply

Great! I agree with you. Otherwise, I could not be sure that our series is convergent. Thank you @Noam Pirani .

Arturo Presa - 5 years, 1 month ago

You can also see that the two subsequences S 3 n + 1 S_{3n+1} and S 3 n + 2 S_{3n+2} have the same limit as S 3 n , S_{3n}, because lim n ( S 3 n + 2 S 3 n ) = 0 \lim_{n\rightarrow \infty} (S_{3n+2}-S_{3n})=0 and lim n ( S 3 n + 1 S 3 n ) = 0. \lim_{n\rightarrow \infty} (S_{3n+1}-S_{3n})=0.

Arturo Presa - 5 years, 1 month ago

Log in to reply

@Arturo Presa Yes, that's even better.

Noam Pirani - 5 years, 1 month ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...