« first day (3028 days earlier)      last day (1988 days later) » 
00:00 - 20:0020:00 - 00:00

8:00 PM
So $f_* : \pi_1(S^1, 1) \to \pi_1(S^1, 1)$ passes down to another homomorphism from $\mathbb{Z} \to \mathbb{Z}$ since $\pi_1(S^1, 1) \cong \mathbb{Z}$ and then that homomorphism is multiplication by some $m \in \mathbb{Z}$ and that $m$ we call the degree of $f$
 
Yup
 
Cool cool, thanks @Dami
 
And $\pi_n(S^n)$ is also $\Bbb Z$ so you have degrees of maps from $S^n$ to $S^n$ as well
Hm, I wonder what happens if you use other homotopy groups of spheres
 
Usually you see degree defined in terms of homology. Now, it turns out that there's this guy called the Hurewicz map, which in this case maps $\pi_1(X) \to H_1(X)$. In this case that guy is an iso so it's all good
 
Yeah, $H_n(S^n)$ is also $\Bbb Z$ and is easier to work with
 
8:03 PM
I was just reading about the Hurewicz map @Dami
 
You know what I need to learn about at some point? Topological dimension
(Small and large inductive dimensions and Lebesgue covering dimension)
 
Is the degree defined in terms of homology more general? Because then we don't need to worry about pointed maps and basepoints and all that jazz
 
It works without basepoints
 
Yeah, this is nice because it generalizes
So, it turns out a closed orientable manifold has top homology $\mathbb{Z}$
 
$\pi_1$ doesn't need basepoints, really, as long as it's connected
 
8:05 PM
@Akiva well, to talk about degree you need to know stuff about induced maps, and I think it's possible that a change of basepoint map can sometimes flip sign. Probably not that big of a deal anyway but maybe
 
The degree is also basically the number of preimages of each point, counted with some sort of multiplicity or something
 
sup nerdos
 
Yeah, that's the other version. If you have a smooth map between closed orientable manifolds, then the preimage of a regular value is gonna be a finite set. If you count each point with either + or - depending on the sign of the determinant of the jacobian, you get the degree.
 
Points could be +2 as well, can't they?
Like $z\mapsto z^2$ on the Riemann sphere
(The suspension of the usual degree-2 map on the circle)
The poles get weird
Unrelatedly
Cuisenaire rods are mathematics learning aids for students that provide an enactive, hands-on way to explore mathematics and learn mathematical concepts, such as the four basic arithmetical operations, working with fractions and finding divisors. In the early 1950s, Caleb Gattegno popularised this set of coloured number rods created by the Belgian primary school teacher Georges Cuisenaire (1891–1975), who called the rods réglettes. According to Gattegno, "Georges Cuisenaire showed in the early fifties that students who had been taught traditionally, and were rated ‘weak’, took huge strides when...
 
@Daminark have ur letter writers gotten to it yet
 
8:13 PM
I haven't really sent in the application forms yet, way too much going on in classes
Probably over Thanksgiving I'll do so
 
schlaggy just sent em in for me
so that's my last round done
 
Excellent
 
highly nervous
 
8:27 PM
Hi chat
 
8:39 PM
Hi. I have $E[|A_n / B_n - C|]$ where $A_n$ and $B_n$ are random variables, and $C$ is a constant. I also know that $B_n\to B=const$ in probability. Any ideas how to get rid of $B_n$ from the denominator by bounding the expression from above?
 
9:01 PM
@AkivaWeinberger Would you mind expanding on that a little bit? The function $f$ is a map from $\Bbb{R}^p$ to $\Bbb{R}^q$. If I'm not mistaken, the mean value theorem doesn't hold for such functions.
 
9:18 PM
Welp, I just posted a question on quadric surfaces, if anyone has time and feels like helping me out. Thanks in advance to any kind hearts and no problem to anyone who doesn't have time or energy for the task!
 
9:36 PM
In case you are interested, I'm trying to solve (d), is it helpful to think about the sequence as $a_n \in \{\frac{m}{n}| 0 \lt m \lt n \}$ ?
 
If I put |f(x)| = x , what are all the possible outcomes for f(x) ? Is it x and -x only or is there an infinite solutions ?
 
depending on whether you want f to be continuous
 
So in general , there is infinite amount ?
 
Yes, your function could equal $x$ or $-x$ on any point and you have infinitely many points to play with, loosely speaking. I'm not sure how to prove it though.
 
9:52 PM
Thanks man , i hope someone answers your question too :)
 
Hmm, not sure if anyone will have time to answer the lengthier description of my question at the moment, but how do I find the reduced equation of a quadratic form given an invertible matrix $Q$ and its diagonal form?
 
I would say that a can be any rational number in [0,1] cuz for any n/m , n<m you will have n/m , 2n/2m , 3n/3m ... desired subsequence.
@FuzzyPixelz
 
Hi :) I have three random variables :1 The result of rolling the dice. 2. The number of reverse(tails) in three coin toss. 3. $\lfloor x^2 \rfloor$ where $x \in [-2,2]$ . Can we describe above random variables on the one, standard probability space? For example $\Omega =[0,1]$
 
10:10 PM
@FuzzyPixelz: It's important to think about the specific ordering of those rational numbers when you're talking about convergent subsequences. It can't just be a random sequence in the first place.
There are lots more limits of subsequences.
 
Yo. I had a thought... There's a more or less straightforward way to assign polynomials to various figures according to their "size", if we assume that a point has "size" 1 and a half-open unit line segment has "size" $s$.
If you do this, an open unit line segment has size $s - 1$, a closed unit line segment has size $s + 1$, a closed unit square has size $s^2 + 2 s + 1$, a closed unit circle... I think that's $\pi s^2 + \pi s$...
 
Hi Ted! How're you? You helped me with a problem I was having yesterday, do you happen to have a moment?
 
Does what I'm talking about ring a bell to anyone?
 
hi @JakeS
 
:) Also, your videos were helpful to help firm up my understanding about some previous theory I hadn't fully digested
 
10:23 PM
ah, cool
I try to be a little bit helpful to my students :P
 
Trying to follow a proof, but hitting a road block. On page 7 here -- www-users.math.umn.edu/~stras087/…, why is the product of the roots * a * (-1)^n = 1?
 
Am I allowed to receive more indications ? @ted
 
@TedShifrin, if I can bother you for just a moment, I've classified a quadratic form, diagonalized it and found an orthonormal basis of eigenvectors. How do I describe its reduced equation?
 
The entries of the diagonal matrix give the coefficients.
@Fuzzy: Can you give me a subsequence converging to $\sqrt2/2$?
 
Euler Lagrange equation (calculus of variatio) is necessay but not sufficient why?
 
10:31 PM
sufficient for what, @henceproved?
 
for minimal
Existence of extremal function
The step in the derivation looked if and only if
 
In differential calculus, does $f'(c)=0$ give you a sufficient condition to guarantee a minimum?
 
oh I get your point, so you mean saddle like situation occurs here
 
or local maximum ... yeah ...
 
What about the lower ordered terms @Ted?
 
10:40 PM
Hi chat
 
You need the change of coordinates from the eigenvectors to transform those.
heya @Astyx
 
Hey guys!
 
hi Demonark
 
I suppose $\frac{1}{n}\left \lfloor n \frac{\sqrt{2}}{2} \right \rfloor \to \frac{\sqrt{2}}{2}$ And we should be able to find our $p$ and $q$ from the enumeration..
But proving things precisely is always a frustration for me
 
OK, @FuzzyPixelz. So what do you think now?
Spivak doesn't intend a pedantic proof on that question. He wants you to understand what's going on.
 
10:44 PM
(have you seen the chatroom I created Ted ?)
 
No @Astyx.
 
Isn't it showing above the starboard on the right ?
 
Zee
@TedShifrin was your PHD advisor chern ?
 
I guess the same trick could work for any number, and we should be able to find a subsequence as long as it's in $[0,1]$
 
@Zee. Yes
 
10:46 PM
But I could really use a pedantic proof
 
Right, @Fuzzy. Any irrational is a limit of a sequence of rationals.
I don't see anything, @Astyx.
 
Huh
 
Zee
@TedShifrin that’s pretty amazing , Chern is one of my fav mathematicians
 
Oh, up at the top, @Astyx. I'm on a desktop.
 
Provided that $\mathbb Q$ is dense in $\mathbb R$, so this a good enough argument ?
 
10:49 PM
Basically. You just need to make sure you choose increasing denominators for your sequence of rationals. Not a big deal.
 
I would feel more comfortable if we could justify that every rational number exists somewhere down that sequence :I
 
The change of coordinates from the eigenvectors - that is, I need to calculate $Q^TB$ where $Q$ is the matrix of eigenvectors and $B$ the coefficients of my lower order terms?
 
Of course, @FuzzyPixelz. He's doing increasing denominators in order ...
Something like that, @JakeS. I can't tell for sure without writing it all out, and I'm not going to do that now.
 
Is it possible to write that down? At this point, I'm somewhat convinced ... I just have this phobia of being asked to describe this more "rigorously" and I have no idea how to do it.
This is very similar to Cantor's diagonal argument (oh yeah... I don't know fancy notation for that either)
 
@TedShifrin is there any corollary to the Fundamental Theorem of Algebra that talks about the product of roots of a polynomial?
 
10:59 PM
Vieta
 
@JoeShmo: This is high school algebra. Just write out the product of $(x-r_j)$ and see what the formula is compared to the original polynomial.
Or you can wiki Vieta formulas.
 
ok, thanks
 
Thanks, Ted! That was certainly a help!
At least I know I'm going in the right direction, haha.
 
I don't remember if I did any examples with linear terms in my lectures, @JakeS. There certainly were a few homework problems with them.
 
So this is a problem on my upcoming rep theory pset. Do you guys have some suggestions for problems to think about?
 
11:08 PM
lol wat
 
Would it make sense for me to find that $x^2 + y^2 -2 z^2 + 2xy -3 \sqrt{2}x +3 \sqrt{2}y = 0$ reduces into something like $-2x^2 + 2y^2 + 3z = 0$ or something like that, where -2, 2, and 0 are my eigenvalues, and $Q^TB$ is (0, 0, 3) where $B$ = $(-3 \sqrt{2}, 3 \sqrt{2}, 0)$?
 
Representation theory is all over quantum mechanics and crystallography.
No, that cannot be right, @JakeS.
Oh, wait. Maybe it can.
You changed $z$ to $x$ in the new coordinate system to confuse me.
 
Wait, I did?
Oohh, hmm.
 
Yeah, it looks reasonable. The $x-y$ linear term corresponds to the $0$ eigenvector.
 
So I would swap my x and z variables around in the reduced form?
 
11:16 PM
No, no, it's fine. It just confused me at first.
 
youtube is down :(
I guess I can't procrastinate then :P
 
Alright! So that's one problem down. Thanks for checking my work Ted.
 
I didn't check very much, @JakeS. Just eyeballed it.
 
Haha, I know, but that's enough for me. Can you eyeball something else?
 
Like what?
 
11:19 PM
I think I have all of the data I need, I just did not know how to organize it. I need to describe the equations of the axes for the reduced form of the quadratic, but I don't know what information that would be, exactly.
 
This is true; do you know if the sorts of questions that use rep theory are ones which can be understood/approached without much physics background?
 
Equations for the axes? Don't they just want the axes as lines in $\Bbb R^3$ (i.e., multiples of the eigenvectors)?
I don't know enough to answer that, Demonark.
 
the only term corresponding to $\alpha_1\alpha_2...\alpha_n$ -- i.e. the only $z$-free term -- is $1$. cute.
 
@TedShifrin I'm not sure. My eigenvectors are -2: (0, 0, 1), 2: (1, 1, 0), 0: (-1, 1, 0), but I don't see how to respond to the question using that information.
Sorry for abusing your help!
 
You should ask your professor what he wants. I told you my interpretation.
 
11:28 PM
wow looks like youtube is having a huge problem
 
y
looks fine to me
 
Will do, Ted! Thank you.
 
Looks fine to me, too, @Leaky. My lectures load fine.
 
some russian fella is having a good time in a dark room in siberia right now
 
11:30 PM
well you can see a huge dot in UK
so maybe that's why
 
yeah. also in LA
and central europe
 
Damn commies everywhere.
3
 
hahahaha
give this man the presidency
 
:0
 
I don't even like being a room owner.
 
11:33 PM
Hello!
 
@TedShifrin well excuse me
 
hi @ÉricoMeloSilva
 
LOL, howdy @Eric.
 
lol hlo
 
Are you here to tell me cool stuff about complex geometry?
 
11:34 PM
Hey @Érico
 
hello @user425181.
 
im here bc im waiting for veggies to roast but also i havent figured out that problem from before
 
which one ... the wrong $\partial\bar\partial$-lemma?
 
I wish I had written myself an answer key to my own exercises years ago, but I sure didn't (and rarely have).
So I have a notation that the proof they gave in Griffiths/Harris uses $d$-exactness rather than the others at the end because they need both $\partial\eta = 0$ and $\bar\partial\eta = 0$ at the end. So can we make up a counterexample?
 
11:41 PM
intuitively it seems to me that d-exact partial-exact and barpartial-exact should imply through the hodge decomposition for kahler manifolds that d-closed forms are orthogonal to the space of harmonic p,q forms -> partialbarpartial-exactness
d-closed + one of those kinds of exactness i mean
cuz the kahler condition means it doesnt matter which operator u consider harmonicity wrt
 
I'm pretty sure that the harmonic forms are the same, of course, but I think the proof uses $d$-exactness to kill the two things I said.
 
Hey everyone
Hey @TedShifrin :)
 
What's an example of a ring theory proof involving a lot of facts that might be accessible to undergraduates? I know this sounds vague, but one where I can look up many different definitions/theorems used, learn them, and come back to it.
 
hmm ok lemme turn off the oven then ill be back here w GH in hand
 
Slap me if I'm being silly.
 
11:51 PM
@Eric: Huybrechts specifically states equivalence assuming $d$-closed.
hi @Perturb
@Ryan: Totally too vague. What do you know?
 
@Ryan how about ED -> PID -> UFD
 
ah HA @Ted there's my issue, i was assuming d-closed as well
 
What is demonic @Alessandro doing awake at this hour?
 
@TedShifrin those are certainly all letters
 
@TedShifrin math
 
11:52 PM
LOL
 
who would have thought
 
i guess now to think about a counter example
 
Cool @Eric
I am sure I had one in mind back in 1980, but that's a long time ago.
 
Nah I'm joking, I was checking the chat before going to sleep in case there's something interesting going on
 
No, we're only being boring, @Alessandro.
 
11:53 PM
@TedShifrin Basics of rings, Euclidean Domains, Principal Ideal Domains and Unique Factorization Domains, polynomial rings.
 
@TedShifrin How are you?
 
see
 
@Ryan: What about quotients?
 
you basically listed them in the order i said
 
@TedShifrin I guess I'll have to go to sleep then :P
 
11:54 PM
doing ok, @user425181.
 
@LeakyNun lol it's the order of my book as well.
 
How about using ring theory and the Gaussian integers to prove that a prime is a sum of squares iff it's $\equiv 1\pmod 4$? @Ryan
 
how about hilbert basis theorem
 
That's one of my favorite proofs in beginning algebra.
Too technical, @Leaky.
 
I mean, it fits his requirements...
how about using galois theory to prove C is alg closed
 
11:56 PM
@TedShifrin Interesting. I'm gonna check that out.
Thanks.
 
@TedShifrin Ithink that's very good, I wolud appreciate it if you helped me solve a multivariable calculus exercise
 
Hilbert basis theorem sounds... professorial @LeakyNun
 
I saw the exercise, @user425181, if you're who I think you are. What specifically do you not know how to do?
 
@TedShifrin Ok, I do not know what I have to do, I think Ii should calculate the rotational of that field and interpret it.
 
00:00 - 20:0020:00 - 00:00

« first day (3028 days earlier)      last day (1988 days later) »