« first day (2774 days earlier)      last day (2543 days later) » 

00:00
Better yet, make a hyperbolic trig substitution to do the integral instead.
Yeah I'm a bit stuck at $e^{\pm cx-cd} = cu+\sqrt{c^2u^2-1}$
You need to use boundary conditions somewhere.
I don't have any
Anyone here like functional analysis?
I have a problem which 'looks like' Banach-Alaoglu but I cant figure it out
lol
But I would recommend using $u=\frac1c\cosh cx$.
Back at that integral.
00:02
I need another hint besides this one
No way to continue from my expression?
Seems so close
Oh, should be $\pm(cx-cd)$ in the exponent
It's very painful to get from the trig to the hyperbolic trig. I don't see it.
part of me also would say once you've written down the integral and done enough integration in your life, resort to mathematica lol
Oh. Here's how you do it. I'm leaving out all the constants. You can put them back.
Sure
00:08
Mathematica isn't the issue. It's hyperbolic trig versus regular trig.
I'm not saying its an issue I'm just saying Mathematica would evaluate the integral for you :)
$e^x = u + \sqrt{u^2-1}$, so $e^x - u = \sqrt{u^2-1}$. Square both sides.
It won't necessarily give the right form of the answer, @Drew.
Omg
I just did that
It'll work, @Lozansky.
I'll try again
Maybe all the constants hindered my success :P
00:09
Yeah, I'm not really sure what the context was @TedShifrin, just a stupid observation from 50 feet above :)
$e^{2x} - 2ue^x + u^2 = u^2 - 1$, so $e^{2x}- 2ue^{x} + 1 = 0$.
Solve for $u$.
@Drew: This is why I teach (taught) hyperbolic trig functions at the beginning of diff geo. That's about the only place people ever use them (plus applied stuff).
The constants are a bit ridiculous, yeah, Lozansky.
Ugh sorry this is taking longer than I thought
@TedShifrin I absolutely hated diff geo the first (and only time) I learned it. It seemed super computational and kinda uninteresting.
But now that I'm a bit wiser, I now know that there are nice connections to geometric analysis, etc. so I'm somewhat motivated to go back.
Depends on the person teaching, as always.
Just don't want to read do carmo again lol
00:12
@TedShifrin Yeah, worked out nicely now :P
@TedShifrin
0
A: What does a nonsymmetric connection on $\Bbb R^n$ look like?

Akiva WeinbergerI asked the question. However, I think I have a simpler example than the currently accepted answer. Let $X=(1,0)$ and $Y=(0,1)$ be the constant unit vector fields in the $x$ and $y$ directions. Define a connection such that: $$\begin{matrix}\nabla_XX=\hat\jmath&\nabla_XY=-\hat\imath\\ \nabla_YX=0&...

Of course, of course. the teacher matters.
You thought the previous discussion was unrelated
Interestingly, a lot of the students who took diff geo (undergrad or grad) from me as undergrads ended up doing PhD's in geometric fields, so I couldn't have been that hateful.
00:13
Awesome! Yeah, do you know who Brian White is
DogAteMy: What the hell are $\vec i$ and $\vec j$. Be consistent!
Yes, I know Brian.
Er, $(1,0)$ and $(0,1)$
I.e., $X$ and $Y$, DogAteMy.
Right, fair point, edited.
anyways @TedShifrin I took a reading course with Brian White on geometric measure theory which was very interesting. at some point I should go back to that
00:15
Finally....
I'm not fond of doCarmo's text for an undergraduate course for various reasons, @Drew. But I don't mind making students do some computations. But even when I TAed for the course at Berkeley, doCarmo was too esoteric in places for some of the good students.
Anyway do I use Morerea's Theorem?
@DrewBrady: You were an undergrad at Stanford?
@TedShifrin I have to chuckle about “applications” as a mere parenthetical
@usukidoll: That's what the hint is telling you to do.
00:16
@TedShifrin Are you talking about the same book I'm using?
Or his surfaces book?
Or if there's another great hint besides this that will be great
No, his undergrad book.
@TedShifrin I am an undergrad at Stanford :)
Amazing
00:16
What really threw my off in my diff geo course was the variational stuff. I always thought "okay, how do we come up with looking at curves that minimize this functional?" (I don't mean geodesics locally minimizing length, that's quite intuitive)
Ooohhhhhhhhh so apply momreras theorem to this
Ah, cool, @Drew. Well, Rafe Mazzeo is a long-time friend and former student of mine.
Er what's a Donald Trump vid doing here?
So all I have to do is apply momore's theorem to the max
00:17
Ah, yeah. Not sure if I'll apply to grad school or not but if I do, then things analytic are in my future lol
And Rafe is a great lecturer
one of my favorites
@Akiva Looking into it
Morning everyone
DogAteMy: Cool. But I like my answer because it justifies my claim that I can make any curve I want to a geodesic!! I'm wondering how yours turns out incomplete (you can't follow geodesics forever). Also, do you know the curvature of your connection?
What do you mean "you can't follow geodesics forever"?
Also, I didn't get up to curvature in do Carmo (Riemannian Geometry) yet
but parallel transport is invariant on the path chosen
00:20
@TedShifrin Also regarding doCarmo, my issue with it was not that it was dense (lol, we read from Rudin last quarter) but rather that it was ... so dry
@Drew: Rafe took differential topology from me as an undergrad and wrote a masters thesis on differential geometry with me (mostly while I was gone).
@Drew: Rudin is way drier, however.
Ah I see. Yeah that's kinda true but now I love Rudin.
I prefer my own notes, for obvious reasons. I don't like having to go through 100 pages of what a $2$-dimensional manifold is to do surfaces, @Drew.
I like the computations but I have no intuition for what's going on.
00:21
Ha. well you spend the first half of an intro course on DG working with curves I feel.
(Incidentally, I love how that's-> still at the top of the starboard)
@BalarkaSen The point is that $\nabla_X(v)=R_{90^\circ}(v)$
Hell no, @Drew. I did only a month (out of 15 weeks).
Also, from what I remember of DG, I remember the computations mainly being helpful to ensure you were digesting definitions. Especially if you had not seen forms prior to the course.
I got through Gauss-Bonnet and a few weeks on hyperbolic geometry.
so that, when I parallel transport a vector along a horizontal line, it just spins
00:22
I don't use forms in my course, since only a small number of the students knew them.
@AkivaWeinberger Hm, I see
I wonder how useful it is to consider manifolds as locally ringed spaces
Ah
This is like what happens when parallel transporting along meridians on the sphere
@BalarkaSen For more intuition, look at the graph of $\ln(\cos(x))$
@BalarkaSen Hm, yeah
Except uniformly everywhere
Yeah
Strange.
00:23
hm. interesting. Is there a recent published D.G. text you'd recommend at the graduate level?
@TedShifrin Which would you recommend, Pugh and/or Rudin?
DogAteMy: Now I'm confused. DO you mean $\ln(|\cos x|)$? At any rate, the geodesic runs off to $-\infty$ when you hit zeroes of $\cos$. I see ...
@Lozansky have you done any analysis before
do you know what it means to say a delta-epsilon proof?
00:23
@TedShifrin It doesn't matter, they have the same components
up to translation
@Lozansky: Rudin is very tough. Pugh expects a lot of maturity, too. There are easier books.
I agree with @TedShifrin
you'll just memorize definitions if you start w/ Rudin +Pugh
Try Abott!
The points is that they're all $\bigcap$s
00:24
(Understanding Analysis)
(I forgot I could do that)
its all on the real line, and you'll wonder if you're missing anything w.r.t. to metric spaces.
The answer is pretty much, no.
DogAteMy: If there's never any holonomy (rotation when you parallel translate around a closed loop), then curvature has to be 0. I want a NON-FLAT example of your phenomenon.
pretty much every metric space can be thought of in a reasonable way as just the real line.
I know you want one, I want one too
Don't know how to make one
00:25
@MatheinBoulomenos According to wikipedia, the hilbert class field $L$ of a number field $K$ is the unique field satisfying $\operatorname{Gal}(L/K) \cong \operatorname{Cl}(K)$. I guess that means that if $K$ is quadratic over $\Bbb Q$ and $L$ is quadratic over $K$ then $L$ is the hilbert class field of $K$....?
so I really encourage you to begin with rigorous analysis on the real line and then only think about generalizing to metric spaces
Rudin goes straight to metric spaces
@ÍgjøgnumMeg that sounds wrong
@DrewBrady I won't have analysis until next year, do you think Abott is something worthwhile to read beforehand just to get a sense of what's going on?
Before I came back into the room, I was doing a computation to see exactly what metric compatibility forces or doesn't force re curvature.
00:26
but I'm not sure why
@Lozansky: I thought you were doing engineering. Now you're doing more math?
Take any quadratic field with class number 2
@MatheinBoulomenos Bleh that's what I meant, not just that $K$ is quadratic lol
If you use the Dirichlet unit theorem and Kummer theory, you can prove easily that there are more than two quadratic extensions of that field
@TedShifrin We have optional courses, among them group/ring theory and analysis
00:27
absolutely, @Lozansky. can't hurt
@TedShifrin I added that comment as a comment (@BalarkaSen)
the Hilbert class field is unique
@MatheinBoulomenos Indeed
I'll still work on it, DogAteMy.
just checking. anyone in this room familiar w/ Banach-Alaoglu?
have a question about it.
@TedShifrin do you have a suggestion for a D.G. book at the intro grad level? something that assumes you know algebra + analysis a
00:30
I know of Banach-Alaoglu
Also rehi Ted, hi Mathein!
There's a grad doCarmo book. If you like differential forms and moving frames, check out Jeanne Clelland's new book. There is a book by Chern and two Chinese guys published by World Scientific (but no exercises).
Depends what you're interested in getting to ...
rehi Demonark
@TedShifrin And the rest are like financial mathematics, regression analysis, numerical methods etc which seem a bit boring
actually, differential geometry would be good for you, Lozansky. Is that on the list?
@TedShifrin one thing I wondered while taking diff geo. Suppose you have a smooth vector bundle $E \to X$ with a connection. Let $H \leq \pi_1(X)$ be the normal subgroup of loops with trivial holonomy. Then parallel transport along loops in $\pi_1(X)/H$ is well-defined, so we get a representation of $\pi_1(X)/H$ with the degree equal to the rank of the bundle. Is this of any use? Can we say anything about the representation in terms of geometric data?
Hi @Daminark
@Ted interesting. Well in my view, a book is useless at this point without exercises. But I will for sure take a look!
00:31
@TedShifrin No, the other math course is discrete mathematics
@Drew: I can send you lots of exercises :P
But look at Clelland for starters.
I've learned weird bits of geometry. For example I did a reading course with Yakov Eliashberg on Riemann Surfaces
I learned algebraic geometry
Masterful guy.
at an intro levle
level
@Balarka will be jealous.
00:32
not trying to brag, but rather to say I feel like there's not really any coherent geometry curriculum!
Say hi to Rafe for me, @Drew. We haven't corresponded in a few weeks.
wow you still keep in touch with him?
I know too little of Eliashberg's works
Yeah, I visit him and his family every year.
Used to go to the math department occasionally, but didn't last visit.
I tried my best to understand whatever I read off of that one book
00:33
(Now that I retired and live in CA. Before that it was less often.)
Wow that's crazy. I don't think any of my undergrad profs will do that for me haha
@Lozansky: Is there a probability course? That is cool stuff.
@Daminark let me type up the problem
@Drew: We also share a strong interest in good food. :)
@Lozansky: I'm not sure I see the point of your doing hard theorem-proving classes unless you're really interested in that. I think a range of good applications might be more fun — probability high on my list. Oh well.
@TedShifrin Yeah, there is an "advanced" probabilty course, with $\sigma$-algebra and stuff
00:36
No, that's not what I meant.
Lozansky that's probably not super useful. I'd recommend a class on Markov chains or something
if you've already seen basic probability
I've already taken the intro level probabilty course
it's a good mix of theory + practice
@Mathein: If $H$ is nontrivial, the bundle has to be flat (i.e., curvature zero).
yeah Markov chains are cool
00:36
Oh, ok.
Have you done a solid linear algebra course with proofs?
@TedShifrin oh right, obviously parallel transport is not homotopy-invariant, duh
Like, we did proofs
On spectral theorem, rank-nullity etc
@Mathein: No, you detect curvature by looking at holonomy on small loops, in fact.
Did you have to write proofs for homework, too, @Lozansky? I don't remember your asking for help with any of that.
I think angle displacement given by parallel transporting along a small loop tends to curvature at that point as the length of the loop tends to 0
Basically the idea is we have two uniformly bounded families {u_n} and {v_n} in a L2 space. Then apparently the following is true: there are weakly convergent subsequences {u_n(k)} and {v_{n(k)}}, a regular borel measure, \lambda a sign function {i.e., mapping to +/- 1} such that u_{n(k)} v_{n(k)} dx -> \lambda d\mu
00:38
Maybe divided by the area bounded by the loop
@Daminark ^
@TedShifrin yeah, that was the whole geometric meaning we did for the curvature tensor
I just remembered as you said that
You can write down the Gauss-Bonnet limit integral, @Balarka. This is also related to Ambrose-Singer in higher dimensions.
Idk how to show that result. The result for weakly convergent subsequences is essentially Banach-Alaoglu. But idk about the signed regular borel measure part.
I removed the appeal to WolframAlpha
00:40
@TedShifrin Some simpler proofs, but mostly just computational exercises
and apparently there's an equivalent form for $x(t)$
Well a class on Markov chains would be a good chance to learn about spectral theory + eigenvalues, etc!
they'll probably review it because a lot of students will come in not knowing what a "spectral decomposition" is .
Proofs in linear algebra is sort of a needed warm-up before doing algebra or analysis (which are all proofs), @Lozansky.
$\cos^{-1}(\operatorname{sech}(t))=\tan^{-1}(\sinh(x))$ (and the latter is better since the former gives the wrong sign for negative $x$)
plus its a good time to learn cute tricks! like you learn to show a linear function is injective it suffices to show the kernel is trivial.
00:42
The parametrization of $\ln\cos$ turned out to just be $x=\int\operatorname{sech}$ and $y=-\int\tanh$, weirdly
So that's quite simple actually
cute little tools like this which will help in higher level courses
There are lots more interesting proofs in linear algebra than such things, @Drew.
We have a tiny course on Markov chains, so I would have to pick another 2-3 courses instead of just 2 in total
@TedShifrin Of course. it depends on the school though.
DogAteMy: This is eerily close to the stuff with the tractrix.
00:43
you may not get a good linear algebra course
Axler is a great book though
good exercises
Most are in fact horrible, but I wrote a good linear algebra book :P
do you like Axler?
I am NOT fond of Axler.
really!
because of his approach to determinants?
We did proofs in linear algebra/calculus/multivariable calculus but the main focus was computations
00:43
@TedShifrin okay, so if we assume that the connection is flat, then my question actually makes sense. Flat connections seem to be non-trivial, since there was a whole workshop once here. It seems there should be a relation between the representation of $\pi_1$ (no need to quotient out $H$, really) and properties of the bundle and/or the connection
I actually found myself referring back to Axler a decent amount when I was in algebra
Yeah, because he does the ring theory to avoid determinants. Determinants are far too important in mathematics (e.g., differential forms).
for example about complexification and things like this
Hah, I figured :)
of course a geometer would not be fond of axler
@Mathein: So flat bundles are constructed from $\pi_1$ representations, of course.
@MatheinBoulomenos A flat bundle is the same as a foliated bundle.
00:44
algebraists love it though because of the 'ring theory' he kind of introduces
Axler is Linear Algebra Done Right, then?
Hey guys. I am struggling to understand a proof (pg. 325 Evans PDE) of the Second Existsnce Theorem for weak solutions. I’m fine with everything up until the statement $\frac{1}{\gamma}(Kf,v)=\frac{1}{\gamma}(f,K^*v)$ where $v$ solves the ellpiptic problem $Lu=f$ in $\Omega$ with $u\vert_{\partial\Omega}=0$ and we have that $Ku=:\gamma L^{-1}_{\gamma}f$ and $h=:L^{-1}_{gamma}f$ and $v=K^*v$ (note that $L_{\gamma}=:Lu+\gamma u$).
@TedShifrin He what? How?
@Drew: I have nothing against algebra, but linear algebra is all about geometry, and that's lost in his book.
00:45
Apparently he has a forthcoming measure theory book, which will probably be quite clear
I consider myself an algebraist and I don't love Linear Algebra Done Right
We consider Mathein an algebraist, too :P
@TedShifrin geometry is one aspect of linear algebra
Sheldon and I were in grad school together, so I know him too.
@TedShifrin you know, I have recently learned that I think my background in geometric linear algebra could be better. things like projection operators as the solution to a variational problem
00:46
Why is that message gathering more stars
Analysis is algebra, as well, it's just that $<$ and $\epsilon$ and $\frac d{dx}$ are elements of the algebraic structure :P
Hiding the algebra for an introductory course (which yours was NOT) as we teach in the US does not help most students.
(Not actually)
Why do you people hate me
I'mma sign off now
00:46
What have I done to you
Nevertheless, most top-selling books delay dot products until the end of the book. Makes me livid.
lol
Balarka I love you
Night, DogAteMy :)
like for example, @TedShifrin do you know this theorem called Tverberg Theorem
Right, yeah, I should sign off
00:47
That's too weird for me to reply to
nope, Drew.
it has a beautiful proof due to Barany which is all linear algebra
@TedShifrin we did almost a whole year of geometric linear algebra in $\Bbb R^3$ and $\Bbb R^2$ in high school. We didn't even define what a vector space is. And dot products were introduced from the beginning
We never used matrices, though
@TedShifrin you may like to read it, it is quite short arxiv.org/abs/1712.06119
00:48
so it was more analytic geometry
Yes, I'm doing that in my advanced high school course right now, Mathein. Your curriculum in Europe is just not comparable to the US. How many times do I have to keep saying that?
@TedShifrin i dont think linear algebra done right is that bad, admit-ably it would be a terrible choice as intro course but as a second take at linear algebra i think his use of abstract algebra is well met.
actually my memory does not serve me it was apparently due to Roudneff, not barany
anyways, if you're interested it is a beautiful proof in convex geometry and linear algebra and is one page. given on page 3 of that link.
@Drew: I downloaded it. Looks like stuff in Babai's lovely combinatorics text (which has tons of linear algebra all through it).
Yeah, I did a reading program with a graduate student last year and we read on the 'local theory of banach spaces' (things like Krien Milman, separating hyperplanes, dvoretsky milman) this is really beautiful stuff, but it made me realize that I could use a lot more experience in geometric linear algebra
00:50
Faust, my bias is quite different. I don't mind teaching linear algebra in an algebra course. That's why I love Artin's algebra book. Do it right.
Well, @Drew, I get teased a lot around here, but my algebra and linear algebra books have the subtitle "A Geometric Approach," so that's obviously my bias.
hm. I should take a look. exercises included ?
The definition of the characteristic polynomial in Axler depends on the fundamental theorem of algebra
Tons. I'm very proud of my exercises. (Same for diff geo and multivariable math, which is multivariable analysis + linear algebra + standard computational stuff.)
that's just wrong
Interesting complaint, @Mathein.
00:52
@TedShifrin i think the problem is virtually no one whos only completed a first course in linear algebra has any idea what the geometric interpretation of what a determinant is
i dont mind avoid ideas that arent intuitive
All the more reason to do it in the second course and not leave it out, @Faust.
avoiding*
It's totally intuitive. Look at my lectures.
im not saying that i dont know what it is
We talked about the geometric meaning of determinants even in my LA course taught by an algebraic number theorist
00:53
just that most people dont understand what a determinant is until much later
That's bad teaching.
even if its explained to them
I agree that most first linear algebra courses are crap.
I loved my first linear algebra course
i hated mine @MatheinBoulomenos giant waste of time im not a engineer
00:55
I'm talking America, @Mathein. You need to understand that every time.
hah I did linear algebra in high school. out of this weird book.
@TedShifrin sorry
Where now our president uses "globalist" as an anti-Semitic adjective. Gotta love the world.
@MatheinBoulomenos the course i took had virtually no content of any use it was an instruction manual for an engineer to get certain number that they needed thats it
"Linear Algebra: An Introduction" by Richard Bronson apparently is what it was called
I actually loved the book. And because it was high school the teaching was actually good
00:56
@Faust I never heard of math majors taking math courses for engineers here in Germany. It sounds annyoing
i dont even think we defined a vector space
The bronson book does!
just row operations and how to computer the determinant to find eigenvalues
I remember trying to memorize the 8 or so axioms lol
seems so silly now. after knowing the right perspective
You can learn serious proofs just staying with $\Bbb R^n$, @Faust, but our linear algebra book does do abstract vector spaces (polynomials, continuous functions, etc.) as well.
00:57
interms of modules over a field
Ted what do you think of the Roman book
@TedShifrin maybe im just slow but i gleaned nothing from my first course in linear algebra
don't know it, Drew.
I already said there are a ton of crap courses out there, Faust.
he has a book on very algebraic linear algebra
like modules over a PID etc.
And lots of horrible teachers.
OK, I'm outta here for now.
peace
00:59
Pieces.
yeaeh I should also get back to work!
thanks for the book ideas Ted :)
if I see Rafe will say hi!
Do that :)
Btw just checking did anyone see my Banach-Alaoglu problem haha
I feel like it just keeps getting lost
I'm looking up cardinal numbers and transfinite induction just so I can be edgiest kid in my measure theory class
01:10
@Drew I'm not fully sure I understand that problem
learning measure theory sounds like a better idea.
@TheTestosteroneFanatic if you're relying on transfinite induction to solve problems in MT something has gone horribly wrong
@PVAL is such a heretic
to be the edgeist kid in MT, open up big rudin and start solving problems from chapter 1.
01:13
To be the edgiest kid in measure theory, use sheaves: andrew.cmu.edu/user/awodey/students/jackson.pdf
"The topos Sh(F ) of sheaves on a σ-algebra F is a natural home for measure theory."
2
RIP
Do you want your professors to cause you physical harm?
@Drew okay wait so I'll TeX it up, so we have $\{u_n\}$ and $\{v_n\}$ are bounded subsets of $L^2(X,m)$. Then you can find some $u_{n_k} \to u$ weakly and $v_{n_k} \to v$ weakly, then some Borel measure $\mu$ and some sign function $\lambda$, such that $u_{n_k}v_{n_k} dm \to \lambda d\mu$
precisely
The statement regarding the weakly convergent sequences should just be a direct consequence of Banach-Alaoglu
Idk what to do regarding the signed regular borel measure part.
also to be clear @Daminark one detail I omitted that is relevant is that the bounded subsets are (a) strongly bounded (i.e., in the L2 norm topology) and (b) uniformly strongly bounded (i.e., the sup of the L2 norm on both families is bounded above), actually by 1 in this case
Overused $\mu$ here
01:17
@MatheinBoulomenos thanks, I'll use anything I find in this in my preliminary exams
i., $\sup_{n} |u_n|_{L^2} \leq 1$. and also for v_n.
also in case it is relevant the space $X = S^1$, the unit sphere of dim 1.
@TheTestosteroneFanatic you have to make remarks like "Wait, you do measure theory, but you don't even define what a localic topos is?"
lol your professor will cry
Where is this from, for reference?
Nah the prof won't cry, he'll just fail you
um, the book we're using is Reed-Simon
but its not from the book
01:24
I see. Well, not quite sure how to approach this, maybe there's some sort of Radon-Nikodym at play in getting the Borel measure? Anyway I'll concede the floor to people who know better than I
Well, Daminark, here's a good start. what does it even mean for $u_{n_k}v_{n_k} dm \to \lambda d\mu$ in $M(X)$?
My guess is that it means the corresponding integrals converge for any integrable fn?
idk though
I was thinking total variation norm?
@TedShifrin Here's a great example, that I didn't realize the significance of when I first saw it:
68
A: What is torsion in differential geometry intuitively?

anonymousHere is an example which I found useful when learning about torsion. Consider $\mathbb{R}^3$. Let $X$, $Y$ and $Z$ be the coordinate vector fields, and take the connection for which $$\begin{matrix} \nabla_X(Y)=Z & \nabla_Y(X)=-Z \\ \nabla_X(Z)=-Y & \nabla_Z(X)=Y \\ \nabla_Y(Z)=X & \nabla_Z(Y)=-...

($\nabla_XX=\nabla_YY=\nabla_ZZ=0$)
That connection is compatible with the metric and has straight lines as geodesics.
Essentially, for $W$ and $V$ constant vector fields, $\nabla_WV=W\times V$.
@Daminark it should be the weak* topology of C_0(X)
And I think it's not flat, either (transport a vector along a square of side length $\pi/2$).
(Or, hell, an equilateral triangle of side length $\pi$. Try transporting a vector normal to the triangle.)
01:36
@Daminark apparently I was right. the notion of convergence we want is vague convergence. so we think of a measure as acting on a continuous function via integration
In mathematics, particularly in the area of functional analysis and topological vector spaces, the vague topology is an example of the weak-* topology which arises in the study of measures on locally compact Hausdorff spaces. Let X be a locally compact Hausdorff space. Let M(X) be the space of complex Radon measures on X, and C0(X)* denote the dual of C0(X), the Banach space of complex continuous functions on X vanishing at infinity equipped with the uniform norm. By the Riesz representation theorem M(X) is isometric to C0(X)*. The isometry maps a measure μ to a linear functional ...
02:10
Hey @KasmirKhaan
Anyone understand topology? I have a hard question
I want to compute $H_2(D^n\times S^1,S^{n-1}\times S^1)$
mathb.in/23095 Banach/Alaoglu can anyone help?
> Suppose $X$ is a locally compact Hausdorff space. Suppose I have two countable families $f_n, g_n \in L^2(X, \mu)$, which are uniformly strongly bounded, i.e.:
$$\sup_{n} |f_n|_{L^2(X, \mu)} \leq 1\\
\sup_{n} |g_n|_{L^2(X, \mu)} \leq 1$$Prove then that there are subsequences $f_{n_k}, g_{n_k}$ which are weakly convergent such that for some regular Borel measure $\nu$ and for a function $\sigma: X \to \{\pm1\}$ which assigns a sign to each element of $X$, it holds that $f_{n_k} g_{n_k} d\mu$ converges to $\sigma \, d\nu$ in the vague topology on $M(X)$.
@DrewBrady You should know, you can turn on LaTeX in chat
Look at the link in the room description (on the top right >>>^^^)
02:29
does anyone here know physics? the physics chat is dead right now...
Press F(ysics) to pay respects
3
I'll ask anyways.. I just need to know what a singular potential is
No idea. What's a potential, even?
Though I have a relevant Wikipedia article
Relevance is the concept of one topic being connected to another topic in a way that makes it useful to consider the second topic when considering the first. The concept of relevance is studied in many different fields, including cognitive sciences, logic, and library and information science. Most fundamentally, however, it is studied in epistemology (the theory of knowledge). Different theories of knowledge have different implications for what is considered relevant and these fundamental views have implications for all other fields as well. == Definition == "Something (A) is relevant to a task...
It is the most relevant Wikipedia article
;(
> To do this work, they used what they called the "Principle of Relevance": namely, the position that any utterance addressed to someone automatically conveys the presumption of its own optimal relevance.
Interesting
02:54
can any one explain where the 2 * 2^k cames from in this : i.imgur.com/MGchu1o.png
@WDUK They're using the fact that $k+1>2$
not sure if i see that clearly
03:33
@WDUK multiply both sides of k+1>2 by 2^k
@WDUK IT IS PURE MAGIC!
more sensibly, if $k$ is a natural number (i.e. a positive integer), then $k+1$ is at least 2, n'est-ce pas?

« first day (2774 days earlier)      last day (2543 days later) »