Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I would like to show that the function $f(x) = (\textrm{cos}2 \pi x, \textrm{sin}2 \pi x)$ is a quotient map; I have already shown that it is surjective and continuous (the latter by invoking the universal property for functions into a product topology (I am considering $S^1$ as a subspace of $\mathbb{R} \times \mathbb{R}$)). However, I'm stuck trying to prove the final condition for $f$ to be a quotient map. Here's my work so far:

Let $U \subset S^1$. Suppose $f^{-1}[U]$ is open in $[0,1]$. Then we need to show that $U$ is open in $S^1$. Let $\bar{y} \in U$; either $\bar{y} = (1,0)$ or $\bar{y} \neq (1,0)$. If $\bar{y} = (1,0)$, then by using the openness of $f^{-1}[U]$, we can deduce that there exists $\epsilon, \epsilon' \in \mathbb{R}_{> 0}$ such that $[0,\epsilon) \cup (\epsilon',1] \subset f^{-1}[U]$ (since $f^{-1}[\{ (1,0) \}] = \{0, 1 \}$). Similarly, if $\bar{y} \neq (1,0)$, there exist $\delta, \delta'\in \mathbb{R}_{> 0}$ such that $(\delta,\delta') \subset f^{-1}[U]$. Of course these sets, $[0,\epsilon) \cup (\epsilon',1]$ and $(\delta,\delta')$ are open in $[0,1]$.

How do I (or can I) proceed from here to show that $U$ is open in $S^1$? On the other hand, I'd also love to hear of some more general methods for showing that the function is a quotient map (maybe certain theorems or something). Thanks in advance for the help!

share|cite|improve this question
up vote 1 down vote accepted

Fix a set $U\subset S^1$ such that $f^{-1}(U)$ is open, and a point $z \in U$. Suppose $z$ has positive second coordinate for the moment. Then there is a unique point in $f^{-1}(z)$ which we'll call $x$, and by the openness of $f^{-1}(U)$ an interval $I=(x-\epsilon, x+\epsilon)\subseteq f^{-1}(U)\cap (0,1/2)$. Consider $f$ restricted to $[x-{\epsilon}/2, x+{\epsilon}/2]$. The function $\cos(2\pi x)$ has a minimum and maximum on this interval, as it is closed and bounded. In fact, by the properties of $I$ above, there are no critical points in the interval so the min and max must occur on the boundary. We will call the min and max $x_0$ and $x_1$. Write $z=(z_0, z_1)$. We claim that, $V=S^1 \cap [(x_0, x_1)\times (0, z_1+1)]$ is an open set containing $z$ and contained in $U$. Clearly it is open by the definition of the subspace topology. By the discussion above, $z_0$ is not the min or max so $x_0<z_0<x_1$, and this gives $z\in V$. Now take $(a, \sqrt{1-a^2})\in V\subseteq S^1$. Since $x_0<a<x_1$, the intermediate value theorem gives that $\cos(2\pi c)=a$ for some $c\in I$. But then $\sin(2\pi c)=\pm \sqrt{1-a^2}$ and it must be the plus sign since $c\in I$. This gives $(a, \sqrt{1-a^2})=f(c)$, and since $c\in I\subseteq f^{-1}(U)$, we have shown that $V\subseteq U$. What we have shown is that $z$ has an open neighborhood around it which is contained in $U$. We want to extend this to all possible $z$, not just ones with positive second coordinate. $Mutatis$ $mutandis$, the same proof goes through for $z$ with second coordinate negative. (E.g., use $(z_1-1, 0)$ in the definition of $V$ instead.) I will leave it to you to modify the details for the remaining two points. You will want to use sine instead of cosine if you do the above min/max argument in these cases. This will show $U$ is open and complete the proof.

share|cite|improve this answer
In your first paragraph, since U in not necessarily connected, you don't know that its preimage is an interval. – Mariano Suárez-Alvarez Nov 2 '13 at 5:17
I didn't claim that the preimage is an interval. Since the preimage is open and $z\neq (1,0)$, the preimage of z must be contained in an open interval which is contained in the preimage of $U$. – Brian Klatt Nov 2 '13 at 5:23
Ah. I read an «is» where there wasn't one :-/ – Mariano Suárez-Alvarez Nov 2 '13 at 5:25
Thanks for the reply, Brian. I'd like to accept this answer because it most directly answers my question. Nice to know that there are more elegant solutions out there, though! – Bachmaninoff Nov 3 '13 at 0:50

You can use the fact that a map $q:X\rightarrow Y $ is a quotient map, if it is continuous, and if it sends saturated open sets to saturated open sets; a subset is saturated if (def.) it containes every fiber {$f^{-1}(y)$} that it intersects.

share|cite|improve this answer
Thank you for replying! We haven't discussed fibers/saturated sets in class. I'll be sure to do some reading on these definitions and the corresponding results. Thanks again! – Bachmaninoff Nov 2 '13 at 2:48
@Bachmaninoff: One can also define saturated set as a set $S$ such that $S=f^{-1}(f(S))$. And, or course, $q$ has to be surjective in order to be a quotient map if it sends saturated open sets to open sets. – Stefan Hamcke Nov 2 '13 at 3:29

Maybe here is a general method:

Note that $[0,1]$ is compact and $S^1$ is Hausdorff. So, actually, your function is a closed map. Therefore, your function is a quotient map.


Let $f:X\to Y$ be a continuous function where $X$ is compact and $Y$ is Hausdorff. Then, $f$ is a closed map.


Pick a closed set $U$ in $X$. Since $X$ is compact, $U$ is compact. Hence, $f(X)$ is compact. Since $Y$ is Hausdorff, $f(X)$ is closed.

share|cite|improve this answer
Thanks for the reply! In lecture, we've yet to touch compactness, but this is very helpful. If you don't mind, could you post the result in a general setting, please? (Would it be something like "if $f: X \rightarrow Y$ is a continuous function from a compact space to a Hausdorff, then $f$ is a closed map"?) – Bachmaninoff Nov 2 '13 at 2:47
@Bachmaninoff I've added a general setting. – Y. Fan Nov 2 '13 at 3:09

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.