Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Question: Show that for any bounded set $E \in \mathbb{R}$, there is a $G_\delta$ set $G$ for which $E \subseteq G$ and $m^*(E)=m^*(G)$.

Let $\{I_n\}$ be a countable collection of open intervals such that $E \subset \bigcup\limits_{n=1}^{\infty} I_n$. Observe the following: $$m^*(E) \leq m^*(\bigcup\limits_{n=1}^{\infty} I_n) \leq \sum\limits_{n=1}^\infty l(I_n) < m^*(E)+\frac{1}{2^n}.$$

Now let $G=\bigcap\bigcup I_n$. $G$ is a $G_\delta$ set. From this, we see that $E \subset G$. Thus for each $n \in \mathbb{N}$: $$m^*(E) \leq m^*(G) \leq m^*(\bigcup\limits_{n=1}^{\infty} I_n) \leq \sum\limits_{n=1}^\infty l(I_n) < m^*(E)+\frac{1}{2^n}.$$

So we take $n \rightarrow \infty$.

Is my proof correct?

share|cite|improve this question
You should denote $I_{n,j}$ the collection of intervals which covers $E$ and approach up to $2^{-n}$ the outer measure of $E$ (which is finite). Here you take $n$ as an index, but we see $n$ outside the sum, and we don't know what it is. But you saw the idea and you just have to take now $G:=\bigcap_{j\geq 1}\bigcup_{n\geq 1}I_{n,j}$. – Davide Giraudo Sep 24 '12 at 7:10
up vote 4 down vote accepted

For each integer $n\geq 1$, let $\{I_{n,j}\}$ be a countable collection of open intervals which covers $E$ such that $$\sum_{j=1}^{+\infty}m^*(I_{n,j})\leq m^*(E)+n^{-1}.$$ Such a collection exists because the outer measure of $E$ is finite, as it's a bounded subset of $\Bbb R$.

Let $G:=\bigcap_{n\geq 1}\bigcup_{j\geq 1}I_{n,j}$. As $\bigcup_{j\geq 1}I_{n,j}$ is open for each $n$, $G$ is a countable intersection of open sets, and it contains $E$. We have for $n\geq 1$, $$m^*(G)\leq m^*\left(\bigcup_{j=1}^{+\infty}I_{n,j}\right)\leq m^**(E)+n^{-1},$$ which concludes the proof.

share|cite|improve this answer
Is the assumption that E is bounded necessary? Royden's definition of outer measure uses inf, so even if the set is not bounded, can't we still find "such a collection"? – hl0202 Feb 2 '15 at 5:35
@hl0202 I was thinking the same thing... – Andrew Sep 9 '15 at 1:49

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.