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.

I have to prove two things.

First is that the Cantor set has a lebesgue measure of 0. If we regard the supersets $C_n$, where $C_0 = [0,1]$, $C_1 = [0,\frac{1}{3}] \cup [\frac{2}{3},1]$ and so on. Each containig interals of length $3^{-n}$ and by construction there are $2^n$ such intervals. The lebesgue measure of each such interval is $\lambda ( [x, x + 3^{-n}]) = 3^{-n}$ therefore the measure of $C_n$ is $\frac{2^n}{3^n} = e^{(\ln(2)-\ln(3)) n }$ which goes to zero with $n \rightarrow \infty$. But does this prove it?

The other thing I have to prove is that the Cantor set is uncountable. I found that I should contruct a surjectiv function to $[0,1]$. But im totaly puzzeld how to do this.

Thanks for help

share|improve this question

2 Answers 2

up vote 5 down vote accepted

Hint: - There is a theorem that if $C_n$ is a descending sequence of measurable sets, $C = \bigcap C_n$ then $\lim_{n \rightarrow \infty}m(C_n)= m(C)$. Here, you know $C_n$ is a descending sequence, $m(C_n) $= (2/3)^n, and you want to know m(C)

  • to prove $C$ is uncountable:

Express a number $x$ between $[0,1]$ in base $3$:

$x =0.x_1x_2x_3...$

In the 1st step, we remove the middle third from $[0,1]$

We express 0=.0, $\frac{1}{3} = .1, \frac{2}{3} = .2$, 1= .222222...

We have 3 intervals: [.0 , .1] , (.1, .2), [.2 , .22222...] and we remove the middle interval. The removed interval (.1 , .2) consists of all numbers with $x_1$ = 1, except the endpoint $.1$ of [.0 , .1 ], however, we can express .1 as .02222.....

So we can use the rule: whenever we have a number of form 0.x1(x is a sequence consisting of 0,1,2) as the end point of an interval, we express as 0.x0222....So in this step, we remove all numbers with $x_1$ = 1. The remaining intervals are [.0, .0222...] and [.2, .222...]

Similarly, we can prove that in the n-th step, we keep only those numbers with $x_n$ = 0 or 2:

So the Cantor set contains of all numbers of the form $.x_1x_2..$ with $x_i =0 $ or 2.

There exists a bijection between $E$ and $[0,1]$: If you consider the new set E, with each member of E is a member of the Cantor set with every digit is divided by 2. E consists of all sequence with $x_i$ is 0 or 1. $|E| = |C|$

There exists an injective map from $[0,1]$ to this new set E. So you can prove it's uncountable.

share|improve this answer
    
Thanks this helps. Do you know a name of the theorem you mentioned at the beginning? Don't think we proved it our lectures. –  Haatschii May 16 '12 at 20:07
    
It's theorem 1.19e in Rudin's Real and Complex Analysis. It's an important theorem, I saw it used in couple of places. It's not hard to prove. the way to prove it is: let $B_1 = C_1, B_2 = C_2\setminus C_1, B_3 = C_3\setminus (C_1 \cup C_2 )$,...hence $B_1,B_2,B_3,...$ are pairwise disjoint, and you can use the property of measure of union of disjoint sets. –  user1412 May 16 '12 at 20:34

You may notice that the Cantor set is homeomorphic (which is even more than required) to $\{0,1\}^{\mathbb{N}}$, which is uncountable. A homeomorphism is established e.g. with $f: \{0,1\}^{\mathbb{N}}\to C$ so that \begin{align*} f((x_{i})_{i=1}^{\infty})=\sum_{i=1}^{\infty}\frac{2x_{i}}{3^{i}}. \end{align*} Since it has a converging geometric serie as a majorant; it definitely converges and even uniformly. Since the components are continuous then $f$ is continuous. You may conclude injectivity by noticing that $|f(x)-f(y)|>0$ if $x\neq y$. Surjectivity is also quite straightforward from the definition since you may produce any element of the cantor set with such a serie by choosing $x_{i}$'s wisely (if $y\in C$, choose $x_{i}=0$ if at $i$:th step $y$ is located in the left part of the $i-1$:th step interval being split to thirds, and $x_{i}=1$ if on the right). Since every continuous bijection from a compact set to a Hausdorff space is homeomorphism, we conclude that $f$ is a homeomorphism.

Since every $C_{n}$ has Lebesgue measure of $(\frac{2}{3})^{n}$ (at each step the measure of the previous step is multiplied with $\frac{2}{3}$) and $\frac{2}{3}<1$, it follows by convergence of measure that \begin{align*} m_{1}(C)=m_{1}(\bigcap_{n=1}^{\infty}C_{n})=\lim_{n\to\infty}m_{1}(C_{n})=\lim_{n\to\infty}(\frac{2}{3})^{n}=0. \end{align*}

share|improve this answer

Your Answer

 
discard

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.