3
$\begingroup$

Let the separable Hilbert cube $Q=\prod_{i=1}^{+\infty}[0,1]$ embed into the real Hilbert space $H=l^2(\mathbb{Z}^+)$, whose coordinate unit vectors are $\{ e_i \}_{i=1}^{+\infty}$, as the subset $\left\{ (x_1,x_2,\dots) \middle| 0 \le x_i \le \dfrac{1}{i^2}\right\}$, and its pseudo-boundary $B(Q)$ embed as $\left\{ (x_1,x_2,\dots) \middle| \exists i \in \mathbb{N}, x_i=0 \vee x_i=\dfrac{1}{i^2}\right\}$. I wonder whether $B(Q)$ is a contractible and locally contractible space.

Since $Q \backslash B(Q)$ is not open in $Q$, $B(Q)$ is not closed in $Q$. Thus $B(Q)$ can not be compact and is not homeomorphic to $Q$, unlike the case between the Hilbert unit ball and unit sphere. There do exist simple closed curves (and other higher dimensional analogues) in $B(Q)$ that are not contained in any finite dimensional subspace of $H$, e.g. the piecewise linear curve $$0 \rightarrow e_1 \rightarrow e_1+\dfrac{1}{2^2}e_2 \rightarrow \cdots \rightarrow \sum_{i=1}^{+\infty}\dfrac{1}{i^2}e_i \rightarrow \sum_{i=2}^{+\infty}\dfrac{1}{i^2}e_i \rightarrow \sum_{i=3}^{+\infty}\dfrac{1}{i^2}e_i \rightarrow \cdots \rightarrow 0$$ which however seems to be null-homotopic since it encloses a "finite area" in $B(Q)$.

$\endgroup$
2
  • 1
    $\begingroup$ @ChristianRemling $B(Q)$ can not be a retract of the punctured Hilbert cube, since any retract of a Hausdorff space is closed, but $B(Q)$ is not closed in the the punctured Hilbert cube (in fact it is a dense subset). $\endgroup$
    – Zerox
    Sep 23 at 17:35
  • $\begingroup$ Indeed. The problem with my map $x\to tx$, $t\ge 1$ chosen as the minimal value that puts $tx$ on the boundary, is that $t$ does not depend continuously on $x$. (Here I realize $Q$ as $[-1,1]^{\mathbb N}$ and delete the origin.) $\endgroup$ Sep 23 at 17:57

0

Your Answer

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

Browse other questions tagged or ask your own question.