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.

Problem. Let $\left \{ a_{n} \right \}_{n\in \mathbb{N}}$ be a sequence where $a_{n}\in \mathbb{N}$, $a_{n}<a_{n+1}$, $a_{n+1}\leqslant 10\cdot a_{n}$ for $\forall n\in \mathbb{N}$. What type of number is $x=0.a_{1}a_{2}a_{3}...a_{n}...$ (concatenation)? (I.e. rational or irrational)

Example: $a_{1}=1, a_{2}=10, a_{3}=100, ... a_{n}=10^{n-1}$ and $x=0.110100...10000...$

share|improve this question
What does $x=0.a_{1}a_{2}a_{3}...a_{n}...$ mean when $a_n$ can be greater than 9, and will certainly be for $n>10$? –  Henry Jan 10 '12 at 14:33
Example $a_{1} = 1, a_{2}=10, a_{3}=100 ... a_{n}=10^{n-1}$ and $x=0.1101001000...100000...$ –  rtybase Jan 10 '12 at 14:38
So concatenation –  Henry Jan 10 '12 at 14:51
What do you know about the decimal expansion of a rational number? Can you find a sequence that produces such an expansion? –  Marc van Leeuwen Jan 10 '12 at 15:00
One way possible way to address the problem is using a result from en.wikipedia.org/wiki/Dirichlet%27s_approximation_theorem which says: α - irrational <=> for ∀ε>0, there ∃ infinite many integers m,n so that 0 < |mα - n| < ε. –  rtybase Jan 10 '12 at 15:05
show 1 more comment

1 Answer

I assume $x=0.a_1 a_2 a_3...a_n...$ just means to concatenate all the $\{a_{n}\}$ to form the decimal expansion of $x$. So the number could be $$0.123456789101112...\qquad(a_{n+1}=a_{n}+1),$$ or $$0.110100100010000...\qquad(a_{n+1}=10a_{n}),$$or whatever. What it can't be is ultimately periodic.

Let $l_{i}$ be the length (in digits) of $a_{i}$. Since $a_{n+1}>a_{n}$, we have $a_{n}\rightarrow\infty$ and hence $l_{n}\rightarrow\infty$; and since $a_{n+1}\le 10a_{n}$, we have $l_{n+1}\le l_{n}+1$. In other words, the sequence of lengths $\{l_n\}$ grows indefinitely without skipping any values. Now suppose $x$ is periodic with period $P$ after the $N$-th digit. Then there must be some $M\ge N+1$ for which $l_{M}$ is a multiple of (say, $k$ times) the periodicity $P$. Hence, $a_M$ falls in the periodic portion of $x$ (since it starts at digit $(\sum_{i=1}^{M-1}l_i) \ge M-1 \ge N$ after the decimal point), and it covers exactly $k$ repeats of the periodic substring. Now, the next $l_{M+1}$ digits of $x$, either $kP$ or $kP+1$, must coincide with $a_{M+1}$. But neither of these cases is possible: the next $kP$ digits are just $a_{M}$ again, and we know that $a_{M+1}>a_{M}$; while the next $kP+1$ digits are $a_{M}$ with a single nonzero digit (the first digit of $a_{M}$) appended, i.e., a number that is strictly greater than $10a_{M}$, and we know that $a_{M+1}\le 10a_{M}$.

We've shown that $x$ can't be ultimately periodic, and hence it must be irrational.

share|improve this answer
mjqxxxx's first example is called the (base-10) Champernowne number. –  J. M. Jan 10 '12 at 15:37
I didn't understood the proof completely, but I got the idea that to prove a number rational, we need to prove some Periodicity into the decimal. And if we prove it is not there, the number is irrational. I know these are very basics, and themathpage.com/aprecalc/… link also helped. Thank you :) –  user840077 Jan 11 '12 at 5:09
add comment

Your Answer


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.