5
$\begingroup$

Let $U$ be an open bounded subset of $\mathbb{R}^n$, and $K$ a compact subset of $U$. Does there always exist a compact subset $L$ of $U$ that contains $K$, and such that $L$ is a retract of $U$.

Looks to me true as I can get as close as I wish with $L$ to the to the limit boundary of $U$ and contain any compact $K$. As $L$ is close enough it will have all the connectivity properties of $U$ and will be a retract... but how do I prove it?

$\endgroup$
4
  • $\begingroup$ Are you making any smoothness assumptions about the boundary of U? $\endgroup$
    – Yemon Choi
    Mar 23, 2017 at 3:13
  • $\begingroup$ Have you looked at the section of Dold's "Lectures on Algebraic Topology" on ENRs? It might give you some ideas of how to prove (or disprove) what you want. $\endgroup$
    – Mark Grant
    Mar 23, 2017 at 11:19
  • $\begingroup$ If you want $L$ to be a subpolyhedron then this is already false for domains in $R^3$: Use $U$ homeomorphic to the Whitehead manifold $W$. (A regular neighborhood of such a retract $L\subset W$ would have to be a PL 3-ball but $W$ cannot be exhausted by 3-balls.) I am not sure about general compact subsets $L$. $\endgroup$
    – Misha
    Mar 23, 2017 at 15:14
  • $\begingroup$ For what it's worth, it is true for $n=2$. $\endgroup$
    – Mizar
    Mar 24, 2017 at 11:26

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.