« first day (2626 days earlier)      last day (2692 days later) » 

19:00
@BalarkaSen I'll agree on "morally equivalent"
Somehow, in complex analysis there are often a lot of ways to show things
@AlessandroCodenotti I heard Galois :D
I actually have a point why one wouldn't want to show FTA with Liouville
how good is transcendence theory?
@Ted I think I was trying to remember the top proof here mathoverflow.net/questions/10535/…
I mean, its potential, among other indicators
19:06
The interesting thing about that proof to me is that they just get the open ness of the map by removing the critical points.
not by using the open mapping theorem.
@MatheiBoulomenos I dunno, I kind of like that proof
What's your argument against it?
@PVAL Oh that is cute
You can use FTA to prove Liouville :D
@Ted There's also an awful comment with lots of upvotes which completely misses the point of the proof.
@LeakyNun what do you mean by "transcendence theory"? Showing that certain constants are trancendent or the theory of transcendence bases etc.
the former
I mean, we still don’t know if $e+\pi$ is rational
19:09
@MatheiBoulomenos Interesting
I think I see what you mean
Actually, I don't prove Liouville itself, but a stronger result: If $f$ is holomorphic and non-constant, then the image of $f$ is dense in $\mathbb{C}$
I need the FTA for that
Right, Casorati-Weierstrass
Ok, I agree
In my eyes it's easier to just prove the little Picard theorem than CW lol
CW is not hard to prove iirc
19:13
I think the standard proof is just to do it on an nbhd around an essential singularity
Can you deduce the open mapping theorem in complex analysis from the fact that submersions are open?
Well locally holomorphic maps are like f(z) = z^k
if you know they have isolated singularities yes.
I think that proves open mapping theorem
@BalarkaSen that's the standard proof, isn't it
19:16
A reparsing of it, sure
by singularities I mean singular points.
Well that follows easily from the fact that the derivative is again holomorphic and the identity theorem
I think
yes the identity theorem though is the non-trivial bit.
That it's like f(z) = z^k upto change of charts really says locally it's a submersion on a deleted neighborhood does it not
Is that how the proof goes?
Well, for $z^k$ it's not difficult to show that is open by working with open balls
19:20
@MatheiBoulomenos Yes, and that's what it needs, really.
But you could argue like that as well
I'm pretty sure the the "it looks like z=z^k" bit comes from expanding f around the point as a power series and approximating it by the smallest order terms.
@PVAL Yes, it's = z^kg(z) where g is biholomorphic
so you change charts and it's exactly like f(z) = z^k in the new coordinates
But when you use the isolated points + submersion argument you don't need to change charts
yeah you prove that by expanding the power series and approximating by the lowest order term.
19:22
@PVAL True, it's just that power series starting with a nonzero constant is invertible
@MatheiBoulomenos Ah...
Ah alright I see what you mean
This is like the argument PVAL linked
sup my dudes
yo @Eric
@Balarka do u wanna do my topology pset for me
19:27
there are some cute probs
email me the pset
it's just hatcher probs
Just type a lot of letters and claim its homotopy equiv. to the correct answers.
lol basically what it'll amount to
19:29
say a lot of handwavish things
one of them is about constructing a complex with $Q_{8}$ as fundamental group
which is nice i guess
oh yeah i remember that
1/2 twist cube bimbadabim done
each face do shit
visualize the 1-skeleton
isn't there a general contruction that works for any group?
19:30
and the 2-cells
just take circles for the generators and disks for relations
@MatheiBoulomenos yeah but he's talking about a specific complex i think?
@MatheiBoulomenos it was a specific construction
yeah sniped
@Eric Anyway I just solved it for you
19:31
i mean i already solved that prob
i did it in three lines
and 8 words
bimbadabim is a nontrivial step man
looool
I also need to do comparison principle stuff for PDE
too much wooooork
But yeah, @Eric, it's kind of a tedious check. I remember trying to wrap my head around that picture.
I don't really like that example, tbh
19:33
The group isomorphism is unsolvable but the homeomorphism problem for 2-complexes isn't.
i liked it but i drew a really dope picture on my bigass whiteboard @Balarka so that helps
Which in some sense means there isn't way to create a list (with repetitions or not) that contained all possible simply-connected 2-complexes.
man i hate hatcher's font tho
@EricSilva Heh
at least not a list where it is decidable if something is an element of that list.
19:36
Can someone please explain this: Suppose there are $n$ points in a plane out of which $m$ points are collinear.Number of triangles that can be formed by joining these $n$ points as vertices=$nC3-mC3$.
For 4-manifolds its conjectured that such a list exists but it is known that the diffeomorphism problem is undecidable.
@Balarka i started reading spanier and then he started talking about categories and i fell asleep
HAH
ur a true homie
like, im not joking, i fell asleep in a cafe
That's how I react to analysis
19:37
lolol
a worker had to come wake me up
it was overall a bad time man
@PVAL-inactive Right, so there's no class of 2-complexes with every possible group appearing as fundamental group of some 2-complex out of that class such that two of them are homeomorphic iff they have isomorphic $\pi_1$, I think?
@MrAP each triangle is uniquely determined by its endpoints, of which there are $n$. Disregarding the $m$ collinear points, there are $\dbinom n 3$ choices. However, the three endpoints can't be all among the collinear points, so $\dbinom m 3$ choices are rejected.
The rest follows from inclusion-exclusion
@EricSilva what do you not like about categories?
@EricSilva Page 408 of Fed
19:40
@Balarka There's no listing of all the simply connected two-complexes, so that for an arbitrary two-complex C there is some computable positive integer N(C) so that if C is simply-connected C is homeomorphic to one of the first N(C) complexes in the list.
@MatheiBoulomenos they're not very useful for the stuff @EricSilva and I do
there's algebraic analysis
The weird thing to me is that there is a conjecture that implies that for 4-manifolds such a list DOES exist.
and that's dreadfully boring
just because something has "analysis" somewhere in it doesn't make analysts automatically interested
and the only reason this doesn't cause a contradiction is the undecidability of the diffeomorphism problem.
19:41
@PVAL-inactive Ah
i mean like, the type of places where category theory shows up in a way that isnt just language is typically not interesting to me
That's quite neat
@EricSilva The result is actually nontrivial. And it's not good enough for the graph result because the resulting $n+1$-current could be something crazy and leave the cylinder. There might be some better result.
It's in Simon too
In my functional analysis class, I proved that the completion of a direct sum is the direct sum of it's completions by using the fact that the completion functor is left-adjoint to the forgetful functor.
@MatheiBoulomenos That's just language though
19:44
@BalarkaSen Well, I didn't have to use the concrete construction of a completion or a direct sum
@BalarkaSen Yeah, I think most places where categories really shine is in more advanced stuff, like categorification.
ah interesting
What is that?
::googles:: Uh, fundamentals?
@TobiasKildetoft I know too little about that, but yeah, probably.
I like the categorical argument that the fundamental group of a topological group is abelian
19:46
I think the first nontrivial application of basic category theory is Yoneda
that's way better than constructing homotopies
@MatheiBoulomenos Meh Eckmann-Hilton can just be seen off concretely
You don't need to construct homotopies
@Balarka my homework begs to differ
It's just a manipulation argument; that's what Eckmann-Hilton says
I think that was on my topology final
19:48
well, but the fact that a functor which preserves products and final objects takes group objects to group objects is trivial
rolleye
I have no idea what any of that means
@Balarka rolls 1 single solitary eye
@EricSilva Ok, so $\Gamma-\Sigma$ bounds an $(n+1)$-current $R$
but does $R$ have to be contained in the cylinder?
@MatheiBoulomenos What sort of algebra do you do by the way?
19:50
I'm just a lowly undergrad, so I'm not sure how to answer this question
I like algebraic number theory
Where do you study
Heidelberg
Oh, right, you're a German dude
@0celo7 i would love to talk about this but i just got home and have to start my algebraic top pset or im gonna have a bad night
and with that, i must be off
19:52
@EricSilva cheers
@TobiasKildetoft next semester I'll be taking courses on algebraic number theory, étale cohomology and group/Galois cohomology
@MatheiBoulomenos That's a lot of cohomology
Bye @EricSilva
i attended a talk on etale cohomology once
But I liked all the algebra I encountered so far, so I'm not sure what sort of algebra I do
19:53
beautiful stuff
I wish I could go to algebra talks
very topologically motivated
not that I would understand anything
Sometimes, when I'm bored (e.g. if my functional analysis prof talks about numerics of PDEs), I just pick a random number, check if there is a simple group of that order on OEIS and if not, try to prove that there is no simple group of that order
2
really fun, I would recommend that for killing time
...
19:55
absolute madman
@MatheiBoulomenos Sounds like you will miss most of the time
@BalarkaSen Just to confirm I am not insane, $H^n(\Omega)=0$ when $\Omega\subset\Bbb R^n$ open?
I mean, at least make sure the number is divisible by either $8$ or $12$ so you have a shot
I don't write down the number itself, but the factorization
to make sure it's an interesting number and not the special case of some theorem
right, so make sure to have enough $2$'s or also a $3$
19:57
why a $3$?
@0celo7 Top cohomology of open $n$-manifolds are trivial, yes.
@TobiasKildetoft @MatheiBoulomenos take $7920 = 2^4 \times 3^2 \times 5 \times 11$
in case the order is not divisible by $8$, since then it is divisible by $3$
(not sure if I know a completely elementary argument for that though)
@LeakyNun There's a simple group of that order, though
the "check OEIS" step is important
also make sure to include at least three primes of course
19:58
of course
@BalarkaSen I'm trying to understand why these guys consider $H^n(\Omega)=0$ to be a restriction...
@0celo7 You need the compactly supported Poincare duality, I believe, to prove this
@MatheiBoulomenos describe the simple group of that order
that's not what I do to kill time :P
@MatheiBoulomenos that sounds like a good way to kill time. I'll do that next time
there's a stat lecture tomorrow
should be feasible
19:59
But beware, some number are just to hard
@BalarkaSen Is there a noncompactly supported PD in this context?
720 is notorious
It's $H^p(M)\cong (H^{n-p}_c)^*$
@0celo7 Yep.
@BalarkaSen Looking at Lee...
20:00
@MatheiBoulomenos You too might find the last of the problems in my exercise set interesting (I don't think Leaky has even really tried starting it yet)
And when $p = n$, $H_c^0$ vanquishes
@TobiasKildetoft what's the problem?
noncompact bro
@MatheiBoulomenos which oeis sequence do you check? I'm still trying to find one
20:01
Leaky has sent me a .pdf with solutions to the rest of them. I might even have time to look at that tomorrow
@TobiasKildetoft yay
@BalarkaSen Lee seems to have a more or less elementary computation of the top cohomologies in all cases
@MatheiBoulomenos thanks
Well (1.) is just the fact that $p$ groups are nilpotent
But it does look like something I enjoy, thanks @TobiasKildetoft
20:05
I should make some progress on my own stuff
what should i do? riemann surfaces, ODEs, galois theory or riemannian geometry?
what is the resolution of this dilemma?
galois theory, without a question
riemannian geometry
finish that damn book
Galois geometry
Covering theory for Riemann surfaces is Galois theory in disguise, so you'll kill two birds with one stone
did I forget to mention algebraist and analysts's opinions don't count?
@MatheiBoulomenos Ah, but that's why I started studying Galois theory
I wanted to make the dictionary clearer
20:08
@MatheiBoulomenos prove that every group of order $1729 = 7 \times 13 \times 19$ is cyclic
@BalarkaSen galois theory
opinions of galois fanbois ddddon't kkkhaount
@LeakyNun Classify numbers such that there are exactly $2$ groups of that order. Then do the same for $3$ groups
@BalarkaSen ::puts on physicist hat:: Um, clearly Riemannian geometry is the most physical and geometrical and pictorial...
@TobiasKildetoft I know that $p^2$ is sufficient
@LeakyNun all Sylow are normal by Sylow, thus it is a direct product of its Sylow, which are cyclic, thus the result follows from CRT
20:11
> product of its Sylow theorems
@MatheiBoulomenos LOL
20:12
> all Sylow are normal
> all Sylow normal by Sylow, thus direct product cyclic Sylow, CRT
Chinese Remainder Sylow
Sylow Sylow Sylow
where is the eject button
I actually learned not too long ago that I have probably been pronouncing the name Sylow incorrectly
say what, 0sylow7?
20:14
saying Sylow this many times cannot be good
@TobiasKildetoft look I don't even know why there is exactly one group of order n iff n and phi(n) are coprime
I think I've been pronouncing Grassmann and Riemann incorrectly.
@BalarkaSen Galois theory is also very geometrical in the sense that it allows us to solve geometric problems which have been open since antiquity
@LeakyNun Hmm, yeah, that might actually involve Burnside's transfer theorem
@PVAL-inactive Oh no, did you say mann like man?
20:16
I'm pretty sure that is the correct way
It's German
and I've been saying it the other way.
Yeah, it's german
Germann
You've been saying it the German way and that's incorrect?
20:16
ya and german has that man sound.
not mon.
@MatheiBoulomenos Echt?
it's "Reemun"
too lazy for IPA right now
are you sure?
I have a german descent professor pronouncing it the other way.
i have always heard in pronounced like that
Wait wait. How exactly @PVAL-inactive
The mann like an American would say man?
20:18
yes
@MatheiBoulomenos $10272 = 2^5 \times 3 \times 107$
Isn't that a sound seems sort of prevelant in German.
Not at all. It doesn't exist.
it is more like an ah
The closest would be män
And that would be ridiculous
20:19
@LeakyNun the $107$ Sylow is normal by Sylow, thus it is not simple
@PVAL-inactive It's not like the mon in Pokemon
The o should be more like an ah
Well my office mate is fluent, so I'll ask him.
I'm fluent
riemon
@0celo7 is right, I'm a native speaker
20:20
Huh
the pokemon go players be craycray for that
I accidentally touched a simple group
that's why you check OEIS first
$33264 = 2^4 \times 3^3 \times 7 \times 11$
@PVAL-inactive When speaking English with a German accent it is sometimes hard to say a German word with a perfect German accent in the middle of speaking English
20:21
checked OEIS
When I am speaking German and say and English word, I say it with an accent, not in perfect American
So how do I pronounce Weyl's first name?
Hermann?
Herm ahn
No, Her-mann
20:22
When I say Cantonese words in English I use the English pronunciations
The a is short
like "dim sum"
sure I just mean the a sound is ah?
No, it's short but not the American short a
like the u in um
say un but more like an a
it's hard to describe
hermoon
20:23
when you say "more like an a"
it makes me think you are interpreting the same sound differently.
Same sound with a different inflection
like as an american
Can someone help me with this: Find the number of words which can be formed by taking two alike and two different letters from the word COMBINATION. I am getting $frac{2P2}{2!}*2P2$.
if I said you pronounce computer like compuder with more of a t than a d.
Describing pronunciation via stack exchange chat is nontrivial
2
20:25
(2P2/2!) *2P2
The pronounciation is compuder though it is weird for us to think of the sound as a d.
@PVAL-inactive Well, like the last a in americano (except slightly shorter I think)
@MrAP COMBINATION -> ABCIIMNNOOT, so we have II/NN/OO/ABCMT; your answer is clearly wrong as I see no 5
and that's the same a in Riemann and Grassmann?
@Tobias?
@TobiasKildetoft yeah that's good
20:26
@PVAL-inactive Close to it at least, but I am not the German speaker here
@PVAL-inactive Tobias' suggestion gets you close enough
I don't know how to say Cartan
The probably correct pronunciation sounds very pretentious
So should it be (3P1/2!) * (5P2)?
The answer is 3C1*7C2*\frac{4!}{2!}$
The answer is $3C1*7C2*\frac{4!}{2!}$
Can someone please explain
?
I am reading a proof and I have a question.

We have that $y-x>0$. By the Archimidean property we have that $n>\frac{1}{y-x}$. There also exists $a,b\in \mathbb{Z}$ such that $a<xn$ and $b>yn$. Why does the last sentence hold? From which property does this hold?
20:50
For the statement "every finite set has no limit point", is it enough to require the underlying space to be T1 space?
i really dont like how they hid this chat even more deeply in the new top bar
what is wrong with this question math.stackexchange.com/questions/2469668/… is it too easy?
already got a close vote and down vote
I used this site for more than a year before I saw the chat. The chat is now the main part that I use. I guess that tells us something about something.
21:54
@PVAL-inactive Oh, god, yeah... I'd friggin love to get my hands on that but it seems to be very rare. No chance of getting it here.
22:34
hey, guys.
hi @LucasHenrique
22:56
@MatheiBoulomenos what letter do you use for an automorphism in Galois theory?
I like to use $\sigma$
good idea
@MatheiBoulomenos how is it that we can't find the closed form of $\sum \frac1{n^2} x^n$ lol
Not sure
I only know that function as an exampe of a power series that diverges everyone on the boundary
well, I mean doesn't absolutely converge anywhere on the boundary
of course it converges for $x=-1$
@LeakyNun isnt that a polylog or some shit
$\psi_2$ and some dorky crap
I see
23:10
@LeakyNun any progress on the $x^6+x^3+1$ problem?
@MatheiBoulomenos thinking
23:43
Does anyone else here hate the new top-nav bar style for .SE sites?
They moved the important stuff from the left to the right.

« first day (2626 days earlier)      last day (2692 days later) »