1
$\begingroup$

I have been trying to solve this question but reaching nowhere

Starting from a countable basis of $\mathbb R$ ,I am asked to construct a Borel set such that $0<m(E \cap I)<m(I)$ for every non empty segment I.

And then must $E$ be of infinite measure?

Here $m$ denotes the Lebesgue measure

$\endgroup$
  • $\begingroup$ such that what? $\endgroup$ – Marios Gretsas Oct 3 at 11:54
2
$\begingroup$

Hint: for a basis, take the intervals $(a,b)$ where $a < b$ are rational.
For your set $E$, take a union of "fat Cantor sets", one for each of these intervals.

$\endgroup$
  • $\begingroup$ I am not allowed to use the Cantor set $\endgroup$ – يمنى الحاج Oct 3 at 12:03
  • $\begingroup$ Why not? Anyway, a "fat Cantor set" is not the Cantor set. $\endgroup$ – Robert Israel Oct 3 at 12:06
  • $\begingroup$ Is it obvious that $m(E\cap I)<m(I)$ for all $I$? @RobertIsrael $\endgroup$ – Kabo Murphy Oct 3 at 12:18
  • 1
    $\begingroup$ @KaviRamaMurthy No, but if you choose the measures of the fat Cantor sets appropriately you can arrange it. $\endgroup$ – Robert Israel Oct 3 at 14:04
0
$\begingroup$

I think the following works; I'll let you worry about verifying that.

Say $(r_j)$ is a dense sequence. Choose $a_j>0$ so that $$a_k>\sum_{j=k+1}^\infty a_j,$$for example $a_j=1/3^j$, then define $$I_j=(r_j-a_j,r_j+a_j),$$ $$E_k=I_k\setminus\bigcup_{j=k+1}^\infty I_j$$and $$E=\bigcup E_k.$$

$\endgroup$

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.