2
$\begingroup$

A perfect set is a closed set with no isolated points. A nowhere dense set is a set whose closure has empty interior. My question is, what is an example of a nonempty perfect nowhere dense subset of $[0,1]$ such that every intersection of the set with an open set is either empty or has positive Lebesgue measure?

Is the Fat Cantor Set an example of such a set? Or does no such set exist?

$\endgroup$
4
$\begingroup$

Any closed set $F\subseteq\mathbb R$ admits a decomposition $F=N\cup F_0$ where $N$ has Lebesgue measure zero and $F_0$ is a closed set whose intersection with every open set is either empty or has positive measure. Of course $F_0$ can have no isolated points. Moreover, if $F$ is nowhere dense then $F_0$ is nowhere dense, and if $F$ has positive measure then $F_0$ is nonempty. Therefore, any nowhere dense closed set $F$ of positive measure will contain a set $F_0$ with the properties you want. Also, if $F$ arises from the usual construction of a "fat Cantor set", then $N=\emptyset$ and $F_0=F$.

P.S. Given a closed set $F$, let $\mathcal U$ be the collection of all open intervals $I$ with rational endpoints such that $F\cap I$ has measure $0$. Then $U=\bigcup\mathcal U$ is an open set and $N=F\cap U$ has measure 0. Let $F_0=F\setminus N=F\setminus U$. Then $F_0$ is a closed set, and the intersection of $F_0$ with any open set, if nonempty, has positive measure.

$\endgroup$
  • $\begingroup$ How do you prove that any closer set has such a decomposition? $\endgroup$ – Keshav Srinivasan Nov 11 '18 at 13:21
  • $\begingroup$ The decomposition can be proved in a way analogous to the "condensation point" method for proving the Cantor-Bendixson theorem. A couple of possibly useful search phrases are "measure dense set" and "metrically dense". $\endgroup$ – Dave L. Renfro Nov 11 '18 at 13:27

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.