« first day (1353 days earlier)      last day (3670 days later) » 

9:24 PM
anybody seen any particularly good questions lately?
 
I just derived this one in Ramanujan's style, but I noted it can be done elementarily without using anything special, but the known alternating series of $\log(2)$ $$\sum_{k=1}^{\infty} \frac{1}{(2 k-1) (16 k^2-16 k +3)}=\frac{\log(2)}{2}$$
It's a nice series!
 
9:44 PM
@Chris'ssis how do you mean in ramanujan's style?
 
@AlexanderGruber I only considered $$\int_0^{\pi/2} \log(\sqrt{\cot(x)}) \cos(x) \ dx = \frac{\log(2)}{2}$$ and I was done.
 
haha, alright.
 
:-)
 
@alexander i saw a dece soft question the other day.
 
9:55 PM
"why do we care about topological groups?" or some such
i like questions about motivation.
 
@Mike so do i, do you know where it is?
 
I answered it and got like 6 votes, so that'd be an easy place to check. Nobody gave a good reason an algebraist should care, so I'd be interested in your thoughts
@alexander lemme know if you answer it
 
@Mike good god the accepted answer is atrocious
i'm editing it for grammar and its taking such a long time
 
Hey guys
"Since the union is disjoint, we must have R_1+R_2=A"
why is that?
am I missing something silly?
 
10:12 PM
@FernandoMartin oh hey, i may know this
 
it's on the third paragraph
 
@seaturtles I couldn't work it out.
 
(of the proof)
 
@FernandoMartin Read Jacobson's take on that.
It's great.
BAII.
 
k
Let's see
 
10:15 PM
so because it's a disjoint union $R_1$ and $R_2$ are closed
 
I don't see why that holds
$R_1$ may not even be prime
 
"If $e$ is an idempotent in $R$, then $X_e$ is open and closed in ${\rm Spec}\, R$, and the map $e\mapsto X_e$ is an isomorphism of the Boolean algebra $E$ of idempotents with $e\times f=ef$, $e\circ f=e+f-ef$ with the Boolean algebra of clopen sets of ${\rm Spec}\, R$."
 
That's the next exercise in AM @Pedro
 
i don't htink you need them to be
 
but then $R_1\not\in \text{Spec}A$
 
10:18 PM
@FernandoMartin Dude.
 
@FernandoMartin we're starting by assuming we're taking $V(R_1)$ and $V(R_2)$ as the disjoint components of $\operatorname{Spec}(A)$ right?
 
Showing $A=R_1+R_2$ amounts to showing $V(R_1+R_2)=\varnothing$.
But $V(R_1+R_2)=V(R_1)\cap V(R_2)=\varnothing$.
Done.
 
That's iiiiiit
Thanks @Pedro
Thank you as well @AlexanderGruber
 
@FernandoMartin Did you watch GoT last Sunday?
 
shocked that i could actually help
 
10:21 PM
I don't watch GoT
 
i sure do hate commutative algebra.
 
Did you read any of the books?
 
@FernandoMartin I cannot see how the algorithm for reducing a matrix to the Smith form works.
I mean.
I cannot prove it works.
 
I don't remember it at all, but what's the problem?
 
10:29 PM
Well, the first part of the algorithm I get it.
Let me explain.
We have a matrix $A=(a_{ij})$, and we want to bring it to a diagonal form, $(d_1,\ldots,d_s,0,\ldots,0)$ with $d_i\mid d_{i+1}$.
Using elementary operations.
Now, suppose first that $A\in D^{m\times n}$, where $D$ is Euclidean, with norm $\delta$.
The first step is to find $a_{ij}$ with minimum norm.
 
By using row and column operations, we may assume $i=j=1$.
Now we divide $a_{1k}$ by $a_{11}$; say $a_{1k}=a_{11}q_{k}+b_{1k}$.
 
Then we do $k$-th column minus $q_k$ times the first.
We have replaced the first row with $b_{1k}$s, with norm less than the previous one.
We repeat this.
 
Exactly
 
10:33 PM
Eventually, things will zero out.
 
Eventually you end up with zeroes
Indeed
Then you do the same with the first column and use induction on the submatrix?
 
No, wait.
 
Well, now we do the same with the first column, that yes.
We end up with a matrix equivalent to $A$ of the form \begin{pmatrix}b_{11}&0&\cdots&0\\ 0&c_{22}&\cdots&c_{2m}\\\vdots&\vdots &\ddots&\vdots\\0&c_{n2}&\cdots&c_{nm}\end{pmatrix}
 
10:37 PM
Well, Jacobson claims we can make it so that the $b_{11}$ (possibly a different one), divides all the elements of the submatrix.
But I don't see how.
 
@alexander I don't like the argument they make at all
 
Let me show what he says @FernandoMartin
 
I can give a lot of nontrivial examples of something I make up that nobody cares about
 
I can read it if you don't feel like typing
 
10:38 PM
That doesn't make it interesting
 
(from the book, that is)
 
"We can also arrange to have $b_{11}\mid c_{kl}$ for every $k,l$. For if $b_{11}\nmid c_{kl}$, we add the $k$th row to the first. Repetition of the first process (picking a minimal pivot, &c) replaces $c_{kl}$ by a nonzero element with a $\delta$ less than that of $b_{11}$. A finite number of steps of the sort indicated will then give a matrix (24) equivalent to $A$ [what I copied above] in which $b_{11}\mid c_{kl}$ for every $k,l$."
@Mike What are you talking about?
 
I think I see it @Pedro
 
@FernandoMartin OK.
I tried to work it out with @seaturtles, but I couldn't.
I thought I did.,
 
Ok, I don't think what I thought works
 
10:45 PM
Let me copy what anon said.
this process you just did decreased $\delta(b_{11})$. in fact, it's guaranteed to do that. what happens when you can't decrease $\delta(b_{11})$ anymore? this process you just did decreased $\delta(b_{11})$. in fact, it's guaranteed to do that. what happens when you can't decrease $\delta(b_{11})$ anymore?
there are only finitely many possibilities for $\delta(b_{11})$, so eventually this algorithm minimizes it
at that point, the negation of the second part of the iff I just stated above becomes true. what is the negation?
the negation is that $b_{11}$ divides all those entries
I think my problem is I didn't continue with the algorithm.
 
Suppose that there's an entry that isn't divisible by $b_11$
 
then you can replace $b_11$ by an element with lower norm
 
@AlexanderGruber In what direction ?
 
Well, that's exactly what anon said
:)
That makes it work IMO
 
10:48 PM
I was studying $$ \ell_k(m) = \int_0^{\pi} \left( \frac{\sin mx}{\sin x} \right)^k\mathrm{d}x $$, but did not find any particularly nice answers for the problem. Haha.
 
If $b_{11}$ does not divide $c_{k,l}$, then you can sum the $k$-th row to the first one and reduce. You end up with an element $b'_{11}$ in your top left corner with $\delta(b'_{11})<\delta(b_{11})$
 
@FernandoMartin Yes.
 
$\delta(b_{11})$ is natural, so you can only do this finitely many times
 
@FernandoMartin Yes.
 
So you'll end up with $b_{11}$ dividing everything after finitely many steps
@PedroTamaroff
 
10:57 PM
@FernandoMartin Why?
 
You can only do the decreasing process finitely many times
if $b_{11}$ didn't divide all entries, you could do it again
So eventually $b_{11}$ does divide all entries - in fact it does after $\delta(b_{11})$ iterations or less
 
Ah.
I wasn't looking at it that way.
 
@Mike i agree
 
3 hours ago, by sea turtles
the algorithm decreases $\delta(b_{11})$ if and only if the remainder of one of the other entries mod $b_{11}$ has $\delta$ smaller than $\delta(b_{11})$.
das wut I sed
 
@seaturtles Right. But I was chasing the $c$s norm, not $b_{11}$s norm. I was thinking about it incorrectly, or too complicatedly.
 
11:05 PM
motivation is important in mathematics. it's easy to study intricate and difficult systems of BS that don't matter at all to anything.
2
 
Now all is good.
@AlexanderGruber What are you guys talking about?
 
motivation > mathematics
 
@PedroTamaroff algebraic geometry.
(just kidding.)
there's a question about why people care about topological groups, and the argument the accepted post made was that they are studied because they come up often enough to have a definition.
 
@AlexanderGruber You said you disslike commutative algebra, but still majored in algebra? What branch ?
 
@FernandoMartin mathematics > anything?
@AlexanderGruber agreed
 
11:07 PM
@pedrotamaroff high five for earlier
 
@N3buchadnezzar I'm in finite group theory and representation theory.
 
@meer2kat Oh?
 
also, i'm not completely against commutative algebra, i am just mad at it right now.
 
@AlexanderGruber Oh, my condolences.
 
never mind, found it
 
11:09 PM
@FernandoMartin here
@Mike I may not answer it. i don't feel comfortable enough with topological groups yet to write an answer i'd be happy with
 
@PedroTamaroff yerp
 
they also arise as group objects over $\text{Top}$
so abstract nonsense is another reason
 
@N3buchadnezzar i'm moving into more applied areas of it lately, design theory and crypto and stuff like that
 
@meer2kat What are we highfiving for?
 
Thats good, moving into matters that matter ;)
 
11:12 PM
@PedroTamaroff for calling the little prick what you did
 
@meer2kat Oh, yes. We need some people to learn their place.
 
I read a bit about elliptic curves modulo n, and their applications in crypto. Fascinating thingies
 
@PedroTamaroff amen. alright gtg again
 
@N3buchadnezzar yeah. even if i'm not the one doing the application, i like knowing my work will someday be good for something
 
hiiiiiiiiiiiiiii
 
11:16 PM
@AlexanderGruber Do you think every mathematician thought that about their work?
 
@N3buchadnezzar definitely not
if there's anything i've learned in math grad school it's that most pure mathematicians are writing papers that nobody is ever going to read, and they know it.
 
@alexander alas. my impression is it's pretty much the weakest thing one can impose on infinite groups to make their representation theory even remotely tractable.
 
@Mike that may be true. you might do well to ask Jim Belk about it if you see him around here or MESE.
 
@AlexanderGruber Is that a bad thing? I mean, someday perhaps somoeone will find some part of it usefull.
 
i love that guy. he has such a great perspective on education.
I'm really glad he's taken such an active role on MESE
 
11:20 PM
@N3buchadnezzar I don't know. it's certainly not how I want my work to be, anyway.
I don't think there's much merit to studying things that are so obscure and esoteric that the main challenge is simply understanding what's even being talked about.
 
Hi @Mike @Alex @N3 @Pedro exhausted
 
hey there @TedShifrin
 
@AlexanderGruber But are not everybody trying to understand "obscure and esoteric" things that others find trivial? Eg what undergrads finds useless majors migth find very usefull in their work.
 
teeedddd
how do I prove that a set is injective and surjective O_O
 
Hi @usukidoll you mean a mapping?
 
11:22 PM
@Ted My algebraic topology professor has taken to jokingly chiding me for not applying to Chicago :P
 
but er I've emailed my prof about this question
http://math.stackexchange.com/questions/751410/prove-f-infty-a-infty-rightarrow-b-infty-is-a-bijection

he said that it's a set, not a map @TedShifrin
 
then he made a typo
 
Yeah, sure @mike
 
there's no such thing as an injective set
@Ted Everyone needs a hobby
 
@Mike after I mentioned to one of my undergrad professors that I got rejected from some school, he asked me why i didn't apply to Princeton, and told me he probably could've pulled some strings if I'd asked him earlier. sooo not what I needed to hear right then. :p
 
11:24 PM
Yeah, I'll need a few in two years :)
 
he gave me three hints..
$f_\infty $ is defined...ex if $A \in A_\infty$ then $f(a) \in B_\infty$
$f_\infty $ is injective ... $f_\infty $ is a restriction of the injective function f
$f_\infty$ is onto... $b \in B_\infty $ there is a $a \in A_\infty$ and $f(a)=b$ element in $A$ maps to $B$ .. belongs to $X_\infty$


All of these take 2 lines each... what the
 
@TedShifrin Grading?
 
@N3buchadnezzar i know what you're saying, but there really is a lot of stuff that's on the fringe, and is just being studied as "maintenance research"
 
Princeton is only for the bravest of souls. I'm exceedingly glad I didn't go there.
 
I thought so... I knew injective sets don't exist but he wrote that whatever I written earlier... didn't make any sense. -_-
 
11:25 PM
@Ted Grading wasn't bad today. Most points were lost from laziness.
 
like, you're in academia, you need to write papers for the pub count, so you take some problems that aren't that hard to solve, but are untouched because they don't really matter.
 
@TedShifrin What do you mean? =)
 
I'm grading tomorrow @Pedro. Most of them are doing the problem on triangulating cylinder and $g$-holed tori ...
 
it's not the main thing in your research, just a stall while you figure harder stuff out
 
@TedShifrin I don't know what that is. =P
 
11:27 PM
@AlexanderGruber Well thats 90% of all Major thesis today so.. What can one do about it ?
 
@N3buchadnezzar shrug nothing. I don't think it's a problem.
 
@Mike @TedShifrin
exact words from my prof when I emailed about the problem....

In the next sentence you talk about $A_\infty$ and
$B_\infty$ being injective and surjective, That does not make sense, $A_\infty$ and $B_\infty$ are sets,
not maps. So what do you need.

1.) Show that $f_\infty$ is defined, i.e., if $a \in A_\infty,$ then $ f(a) \in B_\infty$.
2.) Show that f_\infty is injective, clear because f_\infty is a restriction of the
injective function f.
3.) Show that $ f_\infty$ is onto, i.e.,$if b \in B_\infty,$ then there is an $a \in A_\infty$
 
Well, then you should grade it, @Pedro!
 
@Usukidoll I'm pretty sure the first sentence there is telling YOU there's no such thing as an injective set.
 
@N3: About what?
 
11:28 PM
that just isn't how i would like my research to be, if I need "low hanging apples" I try to get them from applied math so that at least they have a better chance of being used in small real world problems.
 
@AlexanderGruber Well I do not like it either, and I want to do something original that matters. But I clearly see that I have a long way in front of me, and need to dwell into many more useless matters to train to actually at attempt at the problems that do matter
 
@TedShifrin Can you explain?
 
I know that @Mike but this is the exact words my prof wrote to me by email saying that they're sets when come on it's not!!! if they were sets, then clearly $f_\infty : A_\infty \rightarrow B_\infty$ won't even exist
 
Explain what? @Pedro ... Triangulation?
 
@TedShifrin Yes.
 
11:29 PM
but finding unexpected applications of weird pure math subjects has always been what I like to do, so this just may be my personal philosophy
 
D: nerghh
 
@N3buchadnezzar the thing to ask in my opinion is whether or not what you're doing is actually leading to something
 
Struggling for months or years on a single problem must be exchausting and really demotivational as well. I think you need a bit of both.
 
Dividing a manifold (or top space) into subsets homeo to triangles in such a way tgat any pair intersect either emptily, in a single vertex, or in a single edge.
 
@Ted Have I complained about my shampoo yet?
 
11:31 PM
@Mike: Don't.
 
;(
 
@PedroTamaroff it's like you take the manifold and put it into a playstation 1
2
 
what de hell was I assigned?! O_O
 
Hell if I know @Alex
 
You like shampoo too much to see its good name sullied, I see.
 
11:32 PM
ok what if they're maps then we need to show that $f_\infty $ is defined
 
@AlexanderGruber: hahahah, I like that explanation
 
because I think it's bs for an injective set to exist because clearly it's llawlz
 
Hi @Fernando
 
Hi @Ted
 
fernando can you help me ^^
 
11:34 PM
we call him Nando now
 
@TedShifrin What is a triangle in a tori?
 
I don't know what $A_\infty$ is
 
It's a subset homeo to one ... Say in a coordinate patch on the torus.
 
@FernandoMartin It's the subgroup of $S_{\infty}$
 
@Pedro Consider the way you write the torus as a quotient of a square. Draw enough triangles in that square such that none intersects at more than a side
 
11:36 PM
Generated by $3$-cycles.
 
Hahahh
 
I'm not joking!
 
bangs head how can this take 2 lines a piece?!
 
I don't think @usukidoll is talking about groups @Pedro
There's a nice picture of a triangulation of a torus on wiki
 
I'm talking about
http://math.stackexchange.com/questions/751410/prove-f-infty-a-infty-rightarrow-b-infty-is-a-bijection

obviously it's wrong... so I've emailed the prof
and gave me three hints

$f_\infty $ is defined, injective...onto... one sec brb
 
11:37 PM
You didn't define what $A_\infty$ is...
 
@Pedro @Mike: you would have liked my lines about: "Let bi-gons be bygones. And 4-gons are foregone conclusions."
 
@TedShifrin chuckles
 
Thanks for humoring me, @Pedro :)
 
hold up!
@FernandoMartin define what $A_\infty$ is?!
it's a sequence that won't terminate
anyway...


1.) Show that$ f_\infty$ is defined, i.e., if$ a \in A_\infty$, then $f(a) \in B_\infty.$
2.) Show that $ f_\infty$ is injective, clear because $ f_\infty$ is a restriction of the
injective function f.
3.) Show that $f_\infty$ is onto, i.e., if $b \in B_\infty$, then there is an $a \in A_\infty$
and $f(a) = b$. So, show that there is an appropriate element in$ A $ that maps to
b, and then show that it even belong to $X_\infty.$

@FernandoMartin those are the three hints I was given and each of them is only 2 lines
 
@PedroTamaroff here are the triangulations of the torus and klein bottle i have been using lately for practical (computational) purposes
 
11:43 PM
so A is in $A_\infty$... like an A is in the sequence that won't terminate ?!?!!?!?!?! and then the $f(a)$ is in $B_\infty$ ... doesn't terminate and sfjda;lkfjdksl;afj
 
I'm sorry but I can't make sense out of your question
 
@ted can probably fit a pun about "goner" in there somewhere,
 
not even with the hints @fernando
 
Sooner or later @Mike
 
@alex You still in that computational homology class?
 
11:44 PM
@Mike this is the last week.
 
ugh this is worse than the gcd even odd proofs
 
@alexander does everything rely on triangulations?
 
@usukidoll You're trying to prove Schroeder Bernstein?
 
I wanna use that theorem towards that problem @Pedro
 
@Mike well, not exactly
i guess technically nothing relies on triangulations because we don't start with a manifold
 
11:46 PM
Damn it people stop pinging me >:(
 
how much can you tell a computational layman
 
which seems to be the right track, but and i mean buttttttttttt I got toungetied... so the prof gave me these three hints and each one is only 2 lines a piece for the proof which I don't get @pedro
 
@alexbecker but I want to hear about your computational homology class
 
@Mike say i have a manifold and i generate a bunch of data from it
 
what data and how
 
11:47 PM
like, take some randomly selected data points on the manifold
 
did you see the hints @PedroTamaroff ? I'm trying to figure out some of them
 
then i give it to you
 
Sorry @AlexBecker: AlexG's name is way too long.
 
how do you figure out what manifold i made it from?
 
lol
 
11:48 PM
Yes, @AlexG, that's a fascinating problem.
 
I don't know what data points mean. like coordinates of some embedding? values of a real-valued function?
 
@Mike yeah, coordinates of some embedding
 
I guess you're being intentional vague and I should shut up :p
oh ok
 
Why start now? @Mike
 
is that even possible with finitely many coordinates?
 
11:50 PM
Probabilistically speaking, @mike, assuming no tiny holes in it ...
 
like, this is obviously a circle, right?
but how do we tell?
(i'm not asking you, i'm saying, this is the foundational problem of the discipline)
 
as a human being that's obviously a circle, as a contrarian jerk that's obviously the punctured circle aka the real line
 
@Mike well, exactly
 
but I think I get the idea
 
Why not a trefoil or worse knot?
 
11:53 PM
@ted I assume this is embedded in the plane...
 
Oh?:)
 
the basic approach is we build simplicial complexes from the data in various ways, e.g. vietoris ripps complexes with gradually increasing radius, and look at how the betti numbers behave
 
he gave me a picture of a points in the plane you dweeb
also those are homeomorphic, which is I assume what we want?
 
You only thought it was a planar manifold.
 
I think you're due for a knuckle sandwich @Ted
 
11:55 PM
@AlexG: Up to homeo or up to isotopy?
 
@TedShifrin maybe it's bamboo.
@TedShifrin homeo
 
@alexander that's cool. oh shit, I know what you mean now. I heard a talk the other day about this and didn't realize it.
 
@Mike it's big right now
there have been a bunch of applied developments in all kinds of different areas. somebody who recently left our program is using it to detect cancer by finding holes in peoples lungs.
 
Awesome.
 
@TedShifrin i know :) seriously
 
11:57 PM
this was gunnar's talk @Ted
 
ah, cool, @Mike.
 

« first day (1353 days earlier)      last day (3670 days later) »