« first day (3762 days earlier)      last day (1257 days later) » 

12:02 AM
I'm looking for a covering space E of the figure 8, so, two circles that share one point, where the covering projection $p:E -> X$ such that $p^-1(X)$ is not evenly covered. Each point has an open neighbourhood that is evenly covered by the sheets of E, but the whole space X is not evenly covered.
 
0
Q: Let $A$ and $B$ be are non-empty compact sets. If $A\cap B=\varnothing$, then $d(A,B)>0$

AttractorNotStrangeAtAllLet $A$ and $B$ be are non-empty compact sets. If $A\cap B=\varnothing$, then $d(A,B)>0$. We're defining $d(A,B):=\inf\{\left|a-b\right|:a\in A\ \text{e}\ b\in B\}$. I didn't use the fact that they're compact, so I feel something is missing. Is that right? Any feedbacks will be gratefully receive...

Any feedback, if possible.
Also, sorry for posting simple things here, I don't want to bother but at the same time I always learn something
 
12:19 AM
Your choice of k depends on epsilon. Let us write that dependence explicitly as k(epsilon).
Then you know 0 <= k(epsilon) < epsilon.
Your phrasing was, roughly: "I know 0 <= k < epsilon for all epsilon. If k > 0, it follows that k < k/2, by taking epsilon = k/2. This is a contradiction."
But if you encode the dependence on epsilon, all this says is that k(k(epsilon)/2) < k(epsilon)/2, whereas what you wanted was k(epsilon) < k(epsilon)/2.
 
That's right, @Crostul pointed this in the comments as well
Thanks, @MikeMiller. He also gave me a hint on why the sets need to be compact, which I was not using
 
In fact this is false for noncompact sets. I suggest drawing a picture of some in R^2.
Ah
 
@MikeMiller Will do!
 
 
1 hour later…
1:24 AM
it's also false in R already
 
 
2 hours later…
3:04 AM
@Thorgott not for connected closed sets at least :)
 
true
why you can't drop closedness is clear in R already
but I guess seeing why you can't drop boundedness is more natural in R^2
 
3:22 AM
silly question perhaps, but why is $t \ge 0$ in the equality of the indicator functions here -- cpb-us-w2.wpmucdn.com/sites.northeastern.edu/dist/a/42/files/…
 
where exactly in this document
 
$\mathbb{1}_{[0, |f(x)|)}(t) = \mathbb{1}_{\{|f| > t\}}(x)$
The first is $\mathbb{1}_{[0, |f(x)|)}(t) = 1 \iff t \in [0, |f(x)|)$. The second is $\mathbb{1}_{\{|f| > t\}}(x) \iff |f(x)| > t\ (\ge 0\ ?)$.
 
Hello
 
You're asking why this only holds if $t\ge0$?
 
I'm asking where he assumed that $t \ge 0$
 
3:30 AM
I don't see that assumption stated, but it's certainly necessary for that equality to hold
but $t\ge0$ under the relevant integral, so this doesn't turn out to be an issue
 
right
I was just wondering if it was generally true for some ominous reason
 
Someone knows how to evaluate nonperiodic continued fractions?
Well yes I know about Rogers Ramanujan continued fraction but that doesn't help in all cases
Lol right now I am not talking waste...
 
the LHS is $0$ if $t<0$, whereas the RHS is $1$
 
yes, thats what I wrote up there (with a typo)
 
Anyone know a proof that total variation defined on an interval [a,b] via "normal" partitions a<t0<...<t_n<b gives the same result as one where you partition the interval into measurable subsets?
 
4:15 AM
Hello everyone. I'm trying to prove that given two bounded sequences x_n and y_n, and x_n converging to x_0, we have that $\limsup{(x_n-y_n)}\geq x_0 - \limsup{y_n}$
Any hints?
 
 
2 hours later…
6:05 AM
Anyone here?
I have an idea
If I ignore all the users, this room will be mine
Lol just joking
 
 
2 hours later…
7:47 AM
[insert political joke]
 
8:26 AM
if we all ignore each other, we'll all have our own private chatroom
[insert antisocial joke]
 
8:48 AM
I wish there was an ignore button on main. If there was, MSE would be mine.
[I have successfully inserted another antisocial joke]
sorry for my lame jokes
 
9:11 AM
add another plea for forgiveness
 
9:49 AM
@skullpatrol tell me it's relation to math, then I will star it
lol
 
@epic_math Associated with Math.SE; for both general discussion & math questions alike.
 
@EdwardEvans okay
 
10:34 AM
> Rarely if ever expressible as a ratio of integers.
star this message if you prefer "expressable" over "expressible" in the above quotation
 
10:59 AM
where expressable is capable of being expressed
while expressible is able to be expressed.
 
What are you even saying
 
11:11 AM
I think it's an AmE vs BrE difference
 
In other words: incorrect English and correct English
2
 
Okay
Nice
 
11:26 AM
Cant solve it through integrating factors.
Is there any other way?
 
 
2 hours later…
12:58 PM
@user541396 can you include some some source/motivation of your question? Where does it come from? I am interested to know
 
1:20 PM
@epic
It was from a book in the liberary
 
was that a DE book?
 
Yuo
Yup*
 
I don't know much about mathematica but even it can't do this
 
 
2 hours later…
2:54 PM
@user541396 let $u=\arctan(y)$, then you get $$1+\left(x-e^u\right)\frac{\mathrm{d}u}{\mathrm{d}x}=0$$
That simplifies it a bit.
 
$$ \int_{-\infty}^{\infty} \exp\big(-e^x-e^{-x}+x\big) dx $$
any substitution for this?
 
you could try $u=e^x$. That simplifies it a bit, but I don't think it is much easier.
 
I have a naive question. We know that U(n) is path-connected, however, since U(n) is not simply connected that means the continuous deformation of the path from a unitary matrix U to Identity is not homotopic?
 
What do you mean, "is not homotopic"? homotopic to what?
 
3:13 PM
@Astyx you're right, i didn't say it clearly.
 
greetings
 
hi
 
I mean it is not continuously possible to deform the path in the space of unitary matrices to idenity
 
@Astyx hello @Astyx
 
@Shamina That's right, every path is not null-homotopic (homotopic to a constant path)
 
3:16 PM
@robjohn do you think integration by parts after that substitution?
 
hi @Thorgott
 
@geocalc33 The result is a Bessel function, so it may not simplify too much
 
@Astyx Ok, having said that. It means the path connected and continuous deformation are two different aspects?
 
hi
 
hello
 
3:18 PM
Yes. In some sense, path connectedness is saying that all constant paths are homotopic
Simple connectedness is looking at maps from the circle to a space, quotiented by the homotopic relation
 
5
Q: fundamental group of $U(n)$

MarsoIs my logic correct? $f:U(n)\rightarrow U(1)$ defined by $f(A)=\det A$ is a group homomorphism so that the induced homomorphism $f^{*}: \pi_1(U(n))\rightarrow \pi_1(U(1))$ will be an isomorphism, right (I am not sure)? as $\pi_1(U(1))=\mathbb{Z}$ as $U(1)=S^1$ so $\pi_1(U(n))=\mathbb{Z}$.

 
Path connectedness is looking at maps from a point to the space, quotiented by the homotopic relation
More generally you can look at maps from the n-sphere to your space, and connectivity and simple-connectivity are the first two degrees of that aspect
 
@geocalc33 the answer is 0.279732
don't know but probably approximately
 
@Astyx Ok, thanks. Just to understand better, I was referring to this question math.stackexchange.com/a/358554/379191
Then what I understood is that I can diagonalize a unitary matrix by another unitary, thus I can deform it to indemnity, however, this is not a continuous deformation?
 
No, the answer is constructing a continuous path $p:[0,1] \to U(n)$ such that p(0) = I and p(1) = A
 
3:26 PM
@epic_math that's $2K_1(2)$
 
Can someone tell me how to evaluate: $$\int e^{\sin z}dz$$
 
It doesn't mention continuous deformation of paths, which is what I was talking about
 
came up somewhere in my NT research
@geocalc33 oh fine. What about the indefinite integral?
I know that $$\int_{0}^{\pi}e^{\sin z}dz=6.208...$$
 
Are you here? @Balarka
 
Yes but extremely busy lmao
electric field theory
 
3:30 PM
0
Q: Closed form for $I= \int_{-\infty}^{\infty} e^{-e^x-e^{-x}+x}~ dx$

geocalc33Is there a closed form for this integral? $$I= \int_{-\infty}^{\infty} e^{-e^x-e^{-x}+x}~ dx$$ I tried $u=e^x$ but it didn't simplify things much. I think the closed form is a Bessel function, but I can't verify this directly. I'd like to see the steps involved to reach the closed form.

I'm actually kind of interested in how to reach the closed form..
 
@geocalc33 what is the source/motivation of that integral?
 
@BalarkaSen Ah ok, I wanted to ask if you have time to look at Dvoretzky's theorem, maybe another time then!
 
Yeah soon
 
or is that integral completely random :--)
 
@Alessandro The Pineapple Thief is a good band by the way
Plus if you check out later catalogs Gavin Harrison is their drummer
 
3:36 PM
Thanks, I'll check them out
 
The group theory band is the best
 
sigh
 
lol
 
they all use cycles because they are cyclic
the band was created by Abel
and the band is abelian
my next favorite band is the prime numbers
they are mysterious and moody (because their behavior is hard to tell)
 
literally nobody cares
not one single person
 
3:40 PM
not even a half of a person
well leave this
...nobody solved my integral?
not even talked about it?
what about using Taylor expansion of exp?
would that help?
 
@BalarkaSen Have I recommended Dizzy Mizz Lizzy to you before?
 
turning it into an infinite sum will help; I am a god of infinite sums
okay so the integral is $$\sum_{n\ge0}\frac{\int\sin^n(z)dz}{n!}$$
and $$\int\sin^n(z)dz$$ is
I am in action now
is anyone interested?
 
@Astyx Then I see that these two paths are not homotopic to each other? To clear my confusion
 
Homotopy is a concept between paths. Connectivity is a concept between points. Here we are talking about the latter
We have proven that every A is path-connected to the identity, which means that U(n) is path connected
We have not proven that every loop in U(n) is homotopic to a constant loop. That concept is called simple connectedness
 
4:04 PM
@Astyx Thanks a lot! This was a great summary for a confused mind like me ;)
 
Glad to help
 
123
Hi Guys...
Pls share any link or suggest book where i can get complete and intuitive proof of greens, divergence and stoke's Theorem.
 
4:23 PM
Hello everyone. Very quick and stupid question.

For an arbitrary bounded sequence of real numbers $(x_n)_n$, if we define $X_n:=\{x_k:k\geq n\}$, then for every $n$ we can say that $\sup{X_n} \in X_n$, or at least that $\sup{X_n}\in x(\mathbb{N})=X_1$, right?
 
Nope. Take x_n = 1-1/n.
No need to call your questions stupid, that serves no purpose but to make you less confident.
 
123
@user91500 Thanks
Is there book for complete step by step guide.
 
@123 Vector Calculus by Jerrold Marsden and Multivariable Mathematics by Theodore Shifrin are great.
 
123
@user91500 OoKay... Thanks
 
4:39 PM
Ok, @MikeMiller! I'm learning how to properly study something.

Thanks for the counterexample, I'll try to get the good habit of finding examples for what definitions and concepts I'm studying.

I need to show that if $(x_n)_n$ is a bounded sequence of real numbers, then $A=\limsup{x_n}$ is an adherent point of $(x_n)_n$, that is, that there is a subsequence $(x_{n_k})_k$ such that $x_{n_k} \to A$

However, my first take on this was trying to show that $x_{n_k}=\sup{X_k}$. Now I see that's nonsense and that 'adherent point' is a broader concept that allows $A$ to not belong to $x(\mathbb{N})$
 
4:52 PM
Yes, it's important to ground your thinking in examples
Here is another good example. Take $x_n$ to be an enumeration of the rational numbers in $(0,1)$. That means something like the sequence $$(1/2, 1/3, 2/3, 1/4, 3/4, 1/5, 2/5, 3/5, 4/5, 1/6, 5/6, 1/7, \cdots$$
Show that every $x \in [0,1]$ is an adherent point of this sequence. :)
Including $1 = \sup x_n$.
 
 
1 hour later…
5:58 PM
Hi, can I clarify whether the order of $\mathbb{F}_{p^2}^*$ is $p^2-1$ or $p^2-p$?
 
@Thomas Are you confusing this with the set of invertible $2\times 2$ matrices over $\Bbb Z_p$? What are the invertible elements of any field?
 
I think it's more likely he's getting confused with the set of invertible elements in $\Bbb Z/p^2$
 
Integers coprime to field order?
I'm actually confused with this content
 
Note that $\Bbb Z/p^2$ is not $\Bbb F_{p^2}$, since $\Bbb Z/p^2$ is not a field.
@Thomas What is a field?
 
Wait I'll send the link
In modular arithmetic, the integers coprime (relatively prime) to n from the set { 0 , 1 , … , n − 1 } {\displaystyle \{0,1,\dots ,n-1\}} of n non-negative integers form a group under multiplication modulo n, called the multiplicative group of integers modulo n. Equivalently, the elements of this group can be thought of as the congruence classes, also known as residues modulo n, that are coprime to n. Hence another name is the group of primitive residue classes modulo n. In...
 
6:11 PM
Do you know what $\Bbb F_{p^2}$ means?
 
Here in "cyclic case" the order is computed as $p^k-p^{k-1}$
 
Yeah we know all that stuff
You should answer our questions
 
Hmm, may be then I've not understood properly...
Ok,
 
@MikeM: I guess I usually use $R^\times$ for the units of a commutative ring. :P
 
Field is an algebraic structure where addition, subtraction, multiplication and division is possible
 
6:14 PM
That is a terrible definition
 
Always a critic, @Tobias :D
 
Hello Ted,Tobias,Mike
and Thomas
Ted I want to apologize for that day , I really had bad sense of humor , but i am changed now
 
OK, apology accepted. Let's try to keep the room semi-professional. :)
 
Thanks Ted , I won't mind on such small things from now and I will respect others
 
6:55 PM
It suffices to say that $\mathbb{Q}$ is dense in $\mathbb{R}$?

For all $x\in [0,1]$ and for every for every $\varepsilon>0$, we know that there'll be a $n\in\mathbb{N}$ such that $x_n\in(x-\varepsilon,x+\varepsilon)$.
 
A very careful argument requires slightly more work. You need to know that there are infinitely many rationals in $(x - \epsilon, x + \epsilon)$. So if you've already constructed $x_{n_k}$ for $1 \leq k \leq N$, where say $|x - x_{n_k}| < 1/2^k$ (let's say we are converging very quickly), you want to know that there exists a rational in $(x - 1/2^k, x + 1/2^k)$ that you haven't already seen before in the sequence.
 
What is the latex code for restriction? I mean big one for integral
 
\bigg|
\big, \bigg, \biggg, \Big, \Bigg, \Biggg
 
well seems invalid
I'm using overleaf
Oh also need to type |
Thanks
 
To find the MLE of $\theta_2$, could one simply use the fact that the density needs to be normalized, which cancels out $\theta_1$, or is there another approach?
 
7:03 PM
@love_sodam Very confusing question. You mean the vertical bar for evaluating the antiderivative, I take it.
 
yeah I mean vertical bar for evaluation.
 
I tend to use either bigg or Big for that
 
Depends on whether it's in-line or displayed. :)
 
7:51 PM
@TobiasKildetoft Nope, they're good?
 
8:06 PM
Howdy, a @Balarka.
 
Hi @Ted
 
Hi Professor Ted
 
Hi skull.
 
How are you?
 
8:45 PM
How to show $\Sigma_{n=1}^\infty 1-\sqrt{1-2/(6n+1)}$ diverges by comparison test
 
@skullpatrol I've been chuckling about this all morning.
@love_sodam $1-\sqrt{1-\frac2{6n+1}}=\frac{\frac2{6n+1}}{1+\sqrt{1-\frac2{6n+1}}}\ge\frac1{6n+1}$
 
I'm glad you liked it @robjohn :D
 
sqrt(1-x) = 1 - x/2 + O(x^2), so sqrt(1 - 2/(6n+1)) = 1 - 1/(6n+1) + O(1/n^2). Thus your terms are approximately 1/6n. You should try to show that your terms are greater than 1/6n, or maybe to play it safe, just 1/12n.
That would be saying 1 - sqrt(1-2/(6n+1)) > 1/12n, or 1 - 1/12n > sqrt(1-2/(6n+1)), or 1 - 1/6n + 1/144n^2 > 1 - 2/(6n+1), or ...
 
9:06 PM
Oh it works. Thanks
 
9:48 PM
amusing interaction
"This is the fourth time (at least) that you have posted this."
"Then I will ask again, because I think this question is meaningfull "
https://math.stackexchange.com/q/3916997/152424
 
10:07 PM
@s.harp This is a stubborn person who's blunt and perhaps not as smart as he/she thinks he/she is. I've tangled before.
 
10:50 PM
Is the output of a functional derivative itself a function?
 
What do you mean by functional derivative?
 
@TedShifrin I'd say that the ropes used to hang annoying wanna-bes has perhaps advanced math...
 
123
Hi Guys. You are awake? It's 4:00a.m here
 
LOL @robjohn. Annoying presidents too?
3 PM here, @123.
 
11:09 PM
...trumpet has advanced my perception of the power of dirty politics
 
11:48 PM
@MikeMiller I think I know now why our lecturer included an exercise where we needed to show that a chain map induces a map between the direct sums of the homology groups: showing that there is a category where a chain map induces a homomorphism on all homology groups at once, as a "single object"
 

« first day (3762 days earlier)      last day (1257 days later) »