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

Please go to this lecture(1). I have numbered my questions for organization and added the picture which I will constantly refer to. I also apologize in advance for this lengthy question

  1. Right about here, he says that the point (the top one "enclosed" by the semi-circle) isn't open in $X$. I don't understand this. Why exactly isn't that open in $X$

  2. Also he later talks about the semi-circle neighborhood being in $Y$ and that makes the point on the boundary an interior point. This confuses me. He left out the other part of the circle and just ignored that the part of the real "ball" lies outside $Y$ and therefore that ball isn't contained in $Y$.

  3. Also are the sets $X$ and $Y$ even open sets in the lecture? Because later in the follow up video he writes $E = Y \cap G$. How can a metric space (is it even a set?) intersect with a set? Aren't they two different things? Isn't like saying $5 \cap [0,1]$? Sorry if the question is ignorant here.

  4. Later in here, he tries to prove that $E = Y \cap G \iff E$ is open. He says that if a neighbourhood is contained in G, then the intersection of that neighbourhood and $Y$ is a neighbourhood in both $Y$ and $E$. But what happens if he takes a point that's not even in $Y$? And draw a neighbourhood around it such that it doesn't even intersect $Y$ or $E$? And suppose $Y$ really is a set (assuming (3) had been answered and clarifed), How does that imply this mysterious set intersection with an open set gives me and open set? That can't be true in general. What if $(-2,2)\cap [-1,1] = [-1,1]$? That doesn't give me an open set

  5. Finally, I don't understand the idea and strategy in the forward direction of the proof. Like at all...

  6. I found this proof (part of (5)) enter image description here , in one line they say that $d(p,q) < r_p, q \in Y \implies q \in E$. How does that make sense? What if $q$ is outside of $E$?

enter image description here

share|cite|improve this question
up vote 3 down vote accepted
  1. He's saying that the semi-circular set is open in $Y$ but not in $X$. He's not talking about the point itself. The reason this set isn't open in $X$ is because points along the top edge of the rectangle $Y$ have the property that every open ball contains a point in the set $Y$ and a point not in the set $Y$ (draw a little circle around a point lying on that edge, you'll see what I mean).

  2. I'm not entirely sure what you mean. Could you like to a time in the video?

  3. $X$ is a metric space to start with, which is a set "endowed" with a metric. Then $Y$ starts out life as an open set in $X$, which we then 'turn into' a metric space by giving it the restricted metric. As sets, $Y\subset X$. For $E=Y\cap G$, you think of this intersection as plain old ordinary set intersection - $Y$ is still a set even though we've given it a metric.

  4. (and 5.) It's definitely important to note that he's showing $E$ is open in $Y$. For the forward direction, we're actually constructing the set $G$ so that (a) it is open in $X$ and (b) $E=Y\cap G$. You simply construct this by drawing small enough neighborhoods around all the points in $E$. These neighborhoods might 'spill over' into $X$, but that's okay. Then, since neighborhoods are open, we can take an arbitrary union of them (in this case, take the union over all points in $E$) and the result will be an open set. You can then check that (a) and (b) hold. For the reverse direction, it doesn't matter if he picks a point that isn't in $Y$. All we need to show is that each point that is in $E$ has a neighborhood that is contained in $E$. $x\in E$ means that $x\in G$, and $G$ is open so we have our neighborhood. Like he says, this neighborhood intersected with $Y$ is a neighborhood in $Y$, which is what we need.

  5. That such an $r_p$ exists is guaranteed by the assumption that $E$ is open relative to $Y$. I.e. $q$ can't be outside of $E$ because I've said that $d(p,q)$ is small enough so that it isn't. Try drawing a picture.

share|cite|improve this answer
For some reason it's numbering things weird. The last item should be numbered 6. – icurays1 Nov 12 '12 at 20:41
1. That's what I mean, I drew the circle. But the "big" box is $X$, anything interior point in $Y$ is an interior in $X$ no? 2.See edit 3. answered, thank you. I think (1) needs to be clarified before I can understand the others. Thank you – Hawk Nov 12 '12 at 23:34
Since $Y\subset X$ it is true that interior points of $Y$ are also interior points of $X$. However the crux here is thinking about $Y$ in two different ways: as a subset of $X$ and as its very own space. If you look at $Y$ as a subset of $X$, those points along the edge are boundary points of $Y$, not interior points. If you look at $Y$ 'by iteself', they become interior points. Every time you use the word 'interior' you should really add 'with respect to $Y$' or 'with respect to $X$' – icurays1 Nov 12 '12 at 23:46
@icuraysi, let's just say then we do look at $Y$ by itself, (so in our heads, the whole metric X disappears from our picture). Then the set containing a point on the boundary of $Y$ contains points in $Y$ and other things outside our universe, $Y$. Yet this weird ball isn't solely contained in $Y$, it's containing some empty space outside Y. Is the problem I have here that i am not treating $Y$ as my "universe"? – Hawk Nov 12 '12 at 23:49
Correct, any time you look at things 'relative to $Y$', $Y$ becomes your universe. There is nothing outside of $Y$, so thats why these neighborhoods 'on the boundary' look a little funny. – icurays1 Nov 12 '12 at 23:57

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.