« first day (1666 days earlier)      last day (3361 days later) » 

12:17 AM
Could anyone here help me with a linear algebra question?
It's been a long time since I've studied this topic.
 
@Clarinetist OK.
 
So suppose I have a linear transformation $T: V \to V$.
Define $T^k$ to be the composition $\underbrace{T \circ T \circ \cdots T}_{k \text{ times}}$.
Is it true that $T^k$ is a linear transformation as well?
for any $k$?
 
Yes.
Show it is linear.
Can you do this?
 
I can prove that on my own... but that's not the main problem I'm having.
 
Show that if $f,g$ are linear, then so is $fg$.
 
12:20 AM
Someone gave me the following problem:
Suppose $T: V \to V$ is a linear transformation and $\mathbf{v} \in V$ such that "$T^4(\mathbf{v}) = 0$ while $T^{3}(\mathbf{v}) \neq 0$." The question states to prove that $\{\mathbf{v}, T(\mathbf{v}), T^2(\mathbf{v}), T^3(\mathbf{v})\}$ is linearly independent.
So I looked back at a linear algebra book I have.
I know that we must prove that $a_1\mathbf{v} + a_2T(\mathbf{v}) + a_3T^2(\mathbf{v}) + a_4T^3(\mathbf{v}) = 0 \implies a_1 = a_2 = a_3 = a_4 = 0$.
 
I also remember that a set is linearly dependent if any of the elements in it are scalar multiples of each other.
So this is what confused me.
 
Really?
Oh man.
Hmm.
 
A collection of vectors $v_1,\ldots,v_s$ is linearly independent if whenever a sum $\sum a_iv_i=0$, then each $a_i=0$.
Which is what you stated correctly above.
 
12:25 AM
Okay, let's work with that.
 
Now, my recommendation is this:
Apply $T^3$ to your equation.
What gives?
Remember $T^4v=0$.
 
Wow, I couldn't have thought of that idea. So we have $a_1T^3(\mathbf{v}) + a_2T^4(\mathbf{v}) + a_3T^{5}(\mathbf{v}) + a_4T^{6}(\mathbf{v}) = 0$ using linearity.
 
If we suppose $T^3(\mathbf{v}) \neq 0$, then $T^{4}(\mathbf{v}) = 0$.
 
The assumption is that $T^3v\neq 0$ but $T^4v=0$.
 
12:28 AM
Oh, so it's a biconditional?
I wasn't sure how to read "while."
I thought perhaps "A while B" meant if B, then A.
 
No.
It's not a biconditional.
 
So they're just both statements/assumptions.
 
Yes.
It's $A$ and $B$ are true.
 
Oh, okay.
My gut instinct is to look at $T^5(\mathbf{v})$ and $T^6(\mathbf{v})$ next.
So we have $T^5(\mathbf{v}) = T(T^4(\mathbf{v})) = T(0) = 0$?
and I imagine we'll get $T^6(\mathbf{v}) = 0$ similarly since $T^6 = T^2 \circ T^4$ and $T^2$ is also a linear transformation.
 
Yes.
That's fine.
 
12:33 AM
So what does that tell me...
 
Well, doesn't it give you $a_1 T^3v=0$?
 
So that obviously gives $a_1 = 0$.
 
Indeed.
Now continue.
 
Hmm.
Could you clarify why $a_1T^3(\mathbf{v}) = 0$?
 
You got $a_1T^3(\mathbf{v}) + a_2T^4(\mathbf{v}) + a_3T^{5}(\mathbf{v}) + a_4T^{6}(\mathbf{v}) = 0$.
But $T^4v=T^5v=T^6v=0$.
 
12:37 AM
Agreed.
 
So...?
 
So it must follow that since $T^{3}(\mathbf{v}) \neq 0$, $a_1 = 0$. Gotcha.
 
@PedroTamaroff Have you read any Hartshorne?
 
@MikeMiller Just tiny bits of the first chapter.
Why?
 
12:38 AM
Now how to deal with the other $a_i$. Hmm.
 
Same idea, @Clarinetist.
 
What's the definition of an affine open dense subset of a quasi-affine variety?
 
Why are you asking me this?
Ah, wait.
I mixed up Hartshorne and Hatcher.
Hehehehhe.
 
YEESH.
 
@MikeMiller I am doing some exercises in Fulton at the moment.,
 
12:40 AM
So we have $T^4$, $T^5$, and $T^6$, when all evaluated at $\mathbf{v}$, we get $\mathbf{0}$.
 
We have then $(a_2 + a_3 + a_4)\mathbf{0} = \mathbf{0}$.
And this is kinda where I'm stuck. I must be taking things the wrong way.
 
No, no.
Now apply $T^2$.
 
Oh man.
So you're telling me that this whole problem is just this trick of taking linear transformations of linear transformations?
 
Well, it's a useful trick. It is important to understand nilpotent transformations of vector spaces.
It helps you understand all of them in the case your field is algebraically closed. =)
 
12:45 AM
I see that Pedro has made great strides in mathematics.
 
It's going to be a long time before I start reviewing field theory :P
 
A few years ago he did not know how to negate a statement properly.
 
Thank you @Pedro!
 
@ABeautifulMind Hold on now, this is linear algebra.
 
12:59 AM
Hey, anyone know anything about destructive interference..?
I can't seem to figure out why this problem is solved in a way.
 
Disappointed? @TedShifrin
What'd I do? :(
 
@ThomasAndrews I was searching around for topics and I found this one: "Approximate solutions of NP-complete problems".

Do you have any information about that?? Do you think that it is a good topic for a presentation??
 
You could try to describe that millennium problem.
 
If anyone want to help me, I need help with this exercise:
0
Q: Residue of function (two functions multiplied)

JuanI need to know if I'm doing this exercise correctly: $f(z)$ has only one singularity in $z=0$, and it's a pole of order 7 $f(z)=-f(-z)$ $f(z)$ is analitic in $z=\infty$ $g(z)$ is analitic everywhere except of $z=\infty$, where it has a pole of order 7 I need to calculate $Res[f.g]$ in $z=0$ ...

I don't know if Idid it correctly
 
1:15 AM
oh, @Kaj, you turned down being on our team for that problem-solving competition. I dunno if we'll find people.
 
Hello Professor @TedShifrin
 
Oh, I definitely considered it @Ted. But I was telling Dr. Graham that I feel worn thin right now, divided between work for my various courses.
 
I understand, @Kaj, but I was figuring you'd want to do it. He only has one volunteer.
 
I didn't want to commit unless I knew I could invest a nontrivial amount of time.
 
Right. Sarah can't do it, either, because of her Honors thesis.
hi, small one.
 
1:18 AM
That's really unfortunate that there isn't more interest
 
I have a Real Analysis tutoring question...
Claim. If $S, T \subset \mathbb{R}$ are nonempty, then if $S$ is bounded, $S \cap T$ is bounded.
This is trivial when $T$ is bounded.
 
@Juan: I commented on your post. For starters, you're not thinking correctly about how to multiply two series together. Think about multiplying polynomials.
 
$S \cap T \subset S$ ? @Clarinetist
 
OH DUH.
THANK YOU!
 
@Clarinetist: @Kaj beat me to that exact comment.
@Kaj: They've cancelled classes tomorrow at UWG and GaTech. Are we next? SIGH.
 
1:20 AM
So we don't even need to split it in cases!
xP
 
I don't see why you need nonempty, @Clarinetist.
 
No idea. That's just how I was given the problem...
 
The empty set is always bounded.
 
I'm just tutoring right now...
 
Oh, well, remember that a good tutor doesn't do the problems. He makes the student do them :P
 
1:22 AM
Indeed :P
 
@Ted, I got a perfect score on Dr. Graham's complex exam. I have a feeling it would've gone - ahem - differently if it was one of yours though. :P
 
LOL, gee, thanks for the compliment, @Kaj.
Two people got >100 on my diff geo exam. :D
 
That's incredible! Better than last year if I recall correctly
 
I don't berember.
 
I've heard abysmal stories about your homeworks though?
 
1:26 AM
1/3 the class got D's and F's, however. :(
Abysmal? People just not putting in time or getting help? How would that happen?
The 3510 kids are one of the best groups I've ever had, though ... no super-star, but plenty of fantastic students.
 
That's good to hear. Do you know if they've decided on a professor for next year?
 
Yeah, I was involved. Mike Usher will do it.
 
I'm not sure how one "tutors" Microsoft Access online... I just got a request for that.
 
I think any tutoring is hard to do on-line, @Clarinetist, but, unless you can remote load their desktop, it would be nigh impossible.
 
True.
Which reminds me, I have to tutor some actuaries out in Togo (country in Africa) this weekend...
Online.
 
1:30 AM
Actuarial seems doable, as it's just basic probability/stat stuff.
So they can type in how they're approaching the problem, and if Bayes isn't there, you say ...
 
@TedShifrin I answered there :)
 
Analysis, though - judging by my experience in the past few hours... can be difficult if they're 100% lost.
 
Proofs on-line seems hard, although I've coaxed a number of out people here.
 
Do you use LaTeX or something over the internet @Clarinetist ?
 
Ever heard of InstaEDU? Just signed up for it today.
 
1:35 AM
Oops, that was out of, not of out :D
 
I wonder if it supports LaTeX. It would be nice.
 
@Kaj: Did you see my snow alert above?
 
@TedShifrin about UWG and GaTech cancelling class? Indeed. I doubt it's going to snow, however.
 
Wow, why does the east coast seem to be getting hit harder with snow than the Midwest lately? xp
 
LOL, who knows ...
 
1:37 AM
There was no reason for them to cancel class that one Tuesday morning
 
last week? no, indeed, although some people were without power ...
I'm already behind in diff geo; I don't need to lose class on Thursday morning, too.
 
Yeah, last week I guess it was
 
Your good friend Destructicus "didn't make it" to class today. shrug
 
At least it wasn't, what, 3 classes like last year?
 
I think I lost 4 days last year.
 
1:39 AM
He's been fairly sick. Didn't come to the gym with me last night either, and that's a rarity.
 
HARRO.
 
ah, several people have been ...
heya mr @Pedro
 
Some nasty upper respiratory thing that I really, really hope I don't catch
 
@Kaj: THey just announced we're closing at 1:30 tomorrow. Good grief.
Doesn't bode well. Damn.
 
Ugh, that means the gym's going to close as well. Not happy.
 
1:41 AM
Why they do this before they see what weather actually occurs bugs me ...
 
At least I won't miss topology
 
Okay, more linear algebra tutoring... let me see if I get these definitions right.
 
it cancels my office hours which people hardly show up at anyhow. But I'm betting they cancel Thursday now. Evening exam for diff geo coming up ...
 
So if $A$ is a matrix, $\text{Row}(A)$ is the row space of $A$, which is... I suppose, the span of the row vectors of $A$?
 
@TedShifrin How's it going?
 
1:42 AM
don't you hate it when things are logical, @Clarinetist?
 
Lol @Ted
 
Grumpily, @Pedro.
 
Headed to dinner! I'll be back in a bit
 
All that stuff is in my book, of course, @Clarinetist ...
bon appétit, @Kaj
 
Oh dang. I'll have to get to know it better :P
 
1:44 AM
Chapter 4.
 
Morning, @Ted
 
Didn't we already say good night, @Mike?
 
Yes, but I just woke up from a nap.
 
ah, it's almost night naptime for me.
 
Mike, can you help me?
 
1:47 AM
notes he was spurned
 
Probably not, @Pedro, but why not.
 
@MikeMiller So, I have an open connected set $U$ in $S^2$, and a subset $X$ homeomorphic to $[0,1]$ with all but its endpoints inside $U$.
 
Sure.
 
I want to show $U\smallsetminus X$ is connected iff this endpoints are in different components of $S^2\smallsetminus U$.
 
I don't like your symbol. Just use $\setminus$...
 
1:51 AM
Seems right, @Pedro
 
@MikeMiller That's too big.
 
No it's not.
 
Is too.
 
I will not budge on this point.
 
puts both children on ignore
 
1:51 AM
OK.
 
There are two things I attribute any success I've ever had to: volumizing shampoo and \setminus.
 
Well, so I first blindly applied MV to $U$ and $V=S^2-X$. Note $U\cup V=S^2-\{p,q\}$ and $U\cap V=U-X$.
 
Ok.
 
This allows me to relate the ranks of $H_1(U-X)$, $H_1(U)$ and $\widetilde{H}_0(U-X)$.
 
Ok.
 
1:57 AM
surprised @Mike didn't bitch about the oversized tilde
 
Hehe.
 
It doesn't look bad.
 
it looks like Sonny Bono's mustache.
 
He's dead as hell.
 
what's your point?
 
1:58 AM
I would hope his mustache is long gone.
Don't really have one. Didn't think I needed one. :P
 
@MikeMiller Well, here I was tempted to use that $H_1(U)$ is (related to) $H_0(S^2-U)$.
 
Let me think if they're isomorphic.
 
Alexander duality somehow? Surely Fulton hasn't done that much yet.
 
Yeah, no.
He didn't.
But I think he did mention that a subset of $S^2$ is simply connected iff its complement is connected.
Not sure though.
Well, I should use $\widetilde{H}_0$ there then.
 
It should be easier for $S^2$. Anyway, this is true if the complement of $U$ is locally contractible.
And surely he doesn't want to bother making distinctions like that.
 
2:01 AM
In the olden days, that was a definition of simply connected, @Pedro.
Emil Artin's idea, I think.
hears the snores
Update for @Pedro: René just removed the garbage answer, without any hysterics.
 
@TedShifrin Good.
 
I wasn't the only downvote, but ...
 
Downvotes work!
 
Sometimes.
 
Yeah, not for Mhenni.
 
2:14 AM
For example.
I'm hardly around enough to answer or get in fights with people ...
although I rattled Bill's cage last night ...
oh oh, @AlexW is back
 
Hello @Ted. :)
I was just watching a video about Michael Freedman. I feel like anyone would want to be a geometric topologist listening to him.
 
yeah, he was becoming a young star when I was finishing grad school
 
For good reason. He's at microsoft now, studying topological quantum computers.
 
His story is also amazing. If you read his wikipedia entry, you would think it was a smooth (no pun intended) trajectory from one high profile position to the next. But it wasn't like that at all, it seems :)
 
he left academia quite a while ago, didn't he?
 
2:19 AM
I believe so, yes. A decade or so at least.
I read a paper of his (and gompf, morrison, walker) recently, where they proposed some ideas for studying the smooth 4d Poincare conjecture. Shame that the thing they spent most of their computation time on was proved to not possibly work. :P
 
Indeed. I love his anecdote on how he was accepted to Princeton.
 
reminds me of Stallings's paper back in the 60's on how not to prove the (3D) Poincaré conjecture.
 
One of my favorites, @Ted.
His proof and the error is a bit subtle for me, as it stands. But I love the idea of the paper.
 
Well, they didn't know it wasn't how not to prove it yet, @Ted. And the fundamental idea they were doing is still workable. Just the invariant they were using to verify that... doesn't.
It turned out to be just a bit too stable.
 
@Mike: Which Morrison?
 
2:23 AM
Scott. MO moderator.
 
oh, don't know him at all ... I know two other Morrisons, both alg geo (Dave is part physics now at UCSB)
 
He used to be at @Alex's proud alma mater
 
This is by far the longest I have taken to format an answer math.stackexchange.com/a/1164221/33907
 
Before my time, sadly.
 
2:25 AM
Speaking of, I think I give my talk the day after Alex leaves.
 
well, at least Alex will be spared the ordeal ...
 
Nooooooo @Mike. What are you speaking on?
 
Smooth Poincaré conjecture in dimension 4.
 
Quelle coincidence!
 
I guess my job now is to convince you to become a topologist, now that the door's open.
 
2:29 AM
one ex-number theorist converts another?
 
I'm pretty uninitiated, so you might not have too much of a tough job.
 
hehehe @Ted
 
I feel very uncertain. Number theory and arithmetic geometry has a certain aesthetic appeal that's undeniable. But there's so much math I don't know.
 
Me too ...
 
@TedShifrin Whaaaat.
 
2:34 AM
Whaaaat^2?
 
Mike used to be into number theory?
 
Nods.
 
The reason I didn't end up doing arithmetic geometry was because nobody here quite did what I was interested in, @AlexWertheim. Khare is really into modular representation theory, and I'm not. (I was going to say something about Hida but then I remembered I have no comprehension of what he does. I think I tried to say something about how much I liked ____ on my application, but I had honestly no idea.)
 
Interesting @Mike. It's honestly one of my concerns. I don't know enough number theory quite yet to know whether or not my future interests would match up well, and despite being well known, UCLA's number theory department is pretty small, no?
Bill Duke does analytic number theory, which isn't much my cup of tea. It seems most number theory students are studying with Hida or Khare.
 
lots of alg number thy at UGA .... just sayin'
 
2:39 AM
In some sense, sure, but it's vibrant. And yes, that's correct, most are studying with them. You'd have to ask other number theory students - I'm a bad source.
No soliciting, @Ted.
 
As a "turncoat", you might be my best source. =P
 
Is there a sign?
 
Woo, I just made $17.
:D
 
Maybe! I didn't stay in it long enough. Rest assured that people will be honest with you if you ask them questions.
 
@TedShifrin Hehehehe.
 
2:40 AM
@Ted: funny you should say that. I was encouraged to apply to UGA, but later on, and I didn't have the time to get my materials prepared.
 
@AlexWertheim Where are you now?
 
@Pedro: working in Madison, WI. I'll be in grad school next year.
 
@AlexWertheim Ah. Where did you get in?
 
@Pedro: So far (and most likely all): UCLA, UW-Madison, U Penn, Northwestern, and a letter of interest from UNC.
 
@AlexWertheim What's UNC?
 
2:42 AM
I probably would have enjoyed it had I gone that route, but it didn't have the immediate appeal that, say, Poonen's or Bhargava's or Silverman's work did; it felt like they were working on down-to-earth problems with heavy tools, while Hida seems to like studying those technical tools.
 
damn good, @Alex
 
@Pedro: University of North Carolina, Chapel Hill.
@Ted: shucks, you flatter me. I'm fortunate to have excellent mentors.
 
Anyway, you develop your interests in that first year... if you're still not 100% sure what you want to do, you go to a big department that's very active in many different fields... ;)
 
My feelings exactly, @Mike. :D
 
(Also, four number theorists isn't particularly small. The topology group consists of 2 people (we're getting another next year though!); the geometry group consists of 2 people; the combinatorics group consists of 1 person, I think there are five algebraists though they work on a variety of things...)
And once you split analysis into its subsets, the same sort of thing happens.
 
2:49 AM
That's more than fair. I guess I'm not too familiar with what's typical.
 
I think some schools are notable as being very strong in some field or another and pick up a large cloud of people. UCLA is more general, I think, with lots of very strong people - but no particular group really overwhelmingly big.
 
That's been my impression. Both Penn and UW Madison have very strong arithmetic geometry/number theory groups. But UCLA has a very good all around collection of excellent people.
(Which isn't to say its number theory isn't excellent, I just know less about it.)
 
Yeah, Madison's one I know has built a big crowd of such people. I don't know anything about UPenn, but I'll take your word for it :)
 
Penn and UW are both quite broad.
 
I hope visits are informative. I may well feel as you did though, @Mike.
 
2:56 AM
What do you mean by that?
 
You committed to UCLA shortly after visiting, didn't you?
 
Oh, that's what you mean. Yes, I did, but I also at that point didn't have any other options, and was only waiting on two - the visit convinced me that their decisions wouldn't change my mind.
(If you have multiple visit days, I really think you should go to both.)
 
Haha, fair enough. Oh, definitely, I'll be going to all of them, and I imagine I'll wait until the end, even if I do have a strong feeling one way or another.
 
I think that's the right decision.
 
how the hell did I just get the alg top badge? :D
 
3:03 AM
Fundamental group and de Rham cohomology questions... probably.
 
I don't do much with the former.
 
@TedShifrin Well, not by helping me...
>:)
 
I never help you, @Pedro.
@Mike: I enjoyed this one.
 
I always forget the adjunction formula.
 
I have always liked it.
 
3:09 AM
It's very surprising and very nice. And I always forget it.
 
Not hard to prove ...
 
@MikeMiller The set's I've chosen to perform MV are correct.
 
@MikeMiller I have a sequence $0\to H_1(U-X)\to H_1(U)\to H_1(S^2-\{p,q\})\to \widetilde {H_0}(U-X)\to 0$.
 
Go on.
 
3:13 AM
Fulton says I have to see the last thing is not $0$ precisely when the middle map is $0$.
i.e. the map from $H_1(U)$ to $H_1(S^2-\{p,q\})$.
 
mhm
 
The small tilde was sad.
 
I hate the huge space before the subscript.
 
Fixed.
 
I think I would prefer $$\widetilde{H_0(U-X)}$$.
Hmm, it's not making the tilde wide enough.
Ah well. Such are life's tragedies.
 
3:22 AM
@MikeMiller So, I guess I can give an example of what's going on.
Take an equatorial band around $S^2$.
 
I should warn you that I'm not paying attention.
 
Oh, OK.
 
Sorry.
It's easy to hold non-mathematical conversation when you're working on math.
 
I don't understand why I don't have an upvote here
 
@TheEmperorofIceCream I don't think it is really fruitful to discuss that here.
 
3:56 AM
You make it sould like this chat is for fruitfull conversation.
 
4:09 AM
Not really.
 
@MikeMiller It's okay, I wasn't talking ;-)
 
4:48 AM
@Kaj what you learn in top today?
 

« first day (1666 days earlier)      last day (3361 days later) »