16
$\begingroup$

Are there any nice examples of infinite sequences of irrational numbers converging to rational numbers?

One idea I had was the sequence: $ 0.1001000010000001\cdots,0.1101000110000001\cdots,\cdots,0.1111000110000001\cdots,$ etc.

Where the first term in the sequence has ones in the place $i^2$ positions to the right of the decimal point. $(i=1,2,3,\dots)$ For the second term, we keep all the ones from the first term and add one in place of the first zero after the decimal point. We then add ones $i^3$ places after the decimal point. (The logical sum 1+1=1, i.e a number $i^2=j^3$ spaces after the decimal place has the value 1).

It is clear that this will converge to $1/9$, and I don't think the decimal expansion repeats at all.

$\endgroup$
31
$\begingroup$

Pick any irrational number $\alpha$ you like, then consider the sequence $\{x_n=\alpha/n\}_{n=1}^\infty$. Then each term of the sequence $x_n$ is irrational and it converges to zero as $n$ tends to infinity.

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ Wow... Nice answer and it is very easy. +1. $\endgroup$ – Avinash N Oct 6 '18 at 10:31
19
$\begingroup$

How about the sequence $\sqrt{2}/n$, converging to zero?

Or, if you want to see some pattern with the decimal expansion, how about $\frac{\sqrt{2}}{10^n}$ giving 0.141421..., 0.0141421..., 0.00141421... ?

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ Seems I overcomplicated things. Very nice. $\endgroup$ – Daniel Pietrobon Jul 1 '12 at 8:35
13
$\begingroup$

$\large\sqrt[n]{n}$ is irrational when $n>1$, and $\lim\limits_{n\to\infty}\large\sqrt[n]{n}=1$.

See also the related question Existence of irrationals in arbitrary intervals.

$\endgroup$
11
$\begingroup$

Consider the sequence:

$$a_n=\sqrt\frac1{n^2+1}$$

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Although this question just wanted a sequence, does it make sense we search for a faster sequence converging to a rational number? I mean for example, $\sqrt 2/n^4$ is faster decreasing than $\sqrt 2/n$. $\endgroup$ – mrs Jul 1 '12 at 17:19
  • 2
    $\begingroup$ @Babak: You can always find a "faster" sequence. You can take the subsequence whose indices are $n_k=A(k,k)$ where $A$ is the Ackermann function; you can use all sort of non-computable and strange functions to go faster. Or you can just take $a_n=\sqrt[k]\frac1{n^k+1}$... $\endgroup$ – Asaf Karagila Jul 1 '12 at 17:33
  • $\begingroup$ Thanks. I wish I could save this comment alone, not a whole question. :) $\endgroup$ – mrs Jul 1 '12 at 17:41
  • 1
    $\begingroup$ @Babak: Open a notepad/gedit/nano/emacs window, copy and paste the content of the comment. Save on your dropbox/.git/svn server, viola! Alternatively, write a draft email with the content and keep it in a folder for email-notes. $\endgroup$ – Asaf Karagila Jul 1 '12 at 17:50
7
$\begingroup$

What do you mean by nice and better?an example is or right or wrong,there is no other possibility. Put a rational $q$ and take the sequence $q+\sqrt2/n$.

| cite | improve this answer | |
$\endgroup$

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.