4

We know that there does not exist a continuous bijective function from $[0,1]$ to $[0,1]^2$. (More generally, $[0,1]$ is not homeomorphic to $[0,1]^2$.) However, is there a continuous a.e. (almost everywhere) bijective function from $[0,1]$ to $[0,1]^2$?

The motivation for asking this question comes from the following:

1) $([0,1],B_{[0,1]})$ is borel isomorphic to $([0,1]^2,B_{[0,1]^2})$, i.e., there exists a measurable bijective function $f$ from $[0,1]$ to $[0,1]^2$, and its inverse $f^{-1}$ is also measurable.

2) By Lusin's theorem, we have that for every $\varepsilon > 0$, there exists a compact $E ⊂ [0, 1]$ such that $f$ restricted to $E$ is continuous almost everywhere and $\mu (E)>1-\varepsilon $, where $\mu$ is the Lebesgue measure.

So the answer for my question seems to be positive, at least for the case with $[0,1]$ and $[0,1]^2$ replaced by some 1-D space $E$ and 2-D space $E'$. Is there anybody can prove it or disprove it?

Leon
  • 83
  • 1
    You have stated wrongly the definition of homeomorphism. An homeomorphism is a bicontinuous bijective function. – Alessandro Pietro Contini Jun 26 '18 at 09:17
  • @AlessandroContini Thank you for pointing out the mistake! The question has been revised. – Leon Jun 26 '18 at 09:23
  • https://en.wikipedia.org/wiki/Space-filling_curve Maybe this will help a little, though it didn't anwser the question. – Selene Jun 26 '18 at 09:44
  • @XIAODAQU I am looking for a bijective function. The Space-filling curve is only injective. – Leon Jun 26 '18 at 11:29
  • @LeiYu Yeah I'm aware of that. And, do you have any materials about the borel isomorphic between $([0,1],\mathcal{B})$ and $([0,1]^2,\mathcal{B})$? I never heard of this before. – Selene Jun 26 '18 at 12:04
  • @XIAODAQU You can refer to R. M. Dudley's book "REAL ANALYSIS AND PROBABILITY". Or refer to https://math.stackexchange.com/questions/2465128/are-the-measurable-spaces-mathbbrn-bor-mathbbrn-and-mathbbrm – Leon Jun 26 '18 at 12:17
  • @LeiYu ~Thanks! – Selene Jun 26 '18 at 12:27
  • What exactly do you mean with "a.e. bijective"? I'd guess: there is a subset $X\subseteq [0,1]$ of measure 1 on which the function is injective and has measure $1$ as well? – Wojowu Jun 26 '18 at 12:42
  • @Wojowu I mean "a bijective function which is continuous a.e.". – Leon Jun 26 '18 at 12:48
  • I see, I've binded "a.e." to the wrong adjective :) – Wojowu Jun 26 '18 at 12:49

0 Answers0