« first day (1907 days earlier)      last day (3411 days later) » 
01:00 - 19:0019:00 - 00:00

19:00
It might be surjective, that was my first thought, but showing that on the domain side....
Is there an established notation for surjective arrows?
Huy
Huy
@DanielFischer: so if I look at the action via Möbius transformations SL2, for $gz := \frac{az+b}{cz+d}$ I have $g^{-1}z = \frac{dz-b}{-cz+d}$ and $(gz)' = \frac{1}{(cz+d)^2}$, so $(g^{-1}z)' = \frac{1}{(-cz+a)^2}$?
or did you mean something else entirely ._.
19:17
@Huy: What's your goal?
Have any of you guys heard of an "epic morphism" perhaps that what it means?
19:32
@AlecTeal $\twoheadrightarrow$
@AlecTeal An epic morphism/epimorphism is the same thing as a surjective morphism in the category of R-modules.
20:15
Hey @TedShifrin. I took the subject GRE today. Confidently answered 51 of 66 questions, and didn't guess on any of the remaining problems. Not sure if that's good or bad, but I felt decent. Probably could've done a bit better though.
Congrats @KajHansen
Thanks @BalarkaSen
what's the GRE?
What's the subject GRE about?
I'm ready to pass out at this point. That was a long, long exam.
It's a graduate school entrance exam in the USA.
20:17
Exams on what?
@KajHansen Yeah, bet you are exhausted. Have some rest :)
Lots of calculus/analysis, and after that a pretty good smattering of everything else. Combinatorics, probability, algebra, number theory
This one was probably 40% calculus / analysis
@Kaj Long time no see :)
@Huy That, or the corresponding real derivative. Depends on whether $f$ is holomorphic or just smooth.
Hey there @JulianRachman
20:21
How are you?
I'm exhausted right now. I just got done taking a 3-hour really important exam
Well, not "just", but earlier today
well, I wish the best for you
Oh, Connes seems to have a way of reconstructing a Riemannian manifold from it's function space. That's very cool and weird at the same time.
Yeah, I don't feel too bad about it to be honest. Just have to wait a while for my scores :P
20:44
Hope it went well, @Kaj! Got you out of volunteering for the tournament!
@TedShifrin !
I am onto curves right now.
Aha. I went for a 5.5 mike hike up a little mountain and then down to and along a beach, so I haven't looked at your stuff.
That explains why you weren't here for a day.
No. Yesterday I spent with a former student. Today was the hike.
oh, ok.
20:49
good evening @Ted!
well, when you're back and have taken sufficient rest, I'll ask a few things I want to ask you.
(about calc)
Duh, I'm back :)
Hi @Alessandro!
OK. Well, the first question is on gradient. In exercise 3, you have asked to draw the steepest path up towards the mountain top. This can be just done by starting at a point in the bottom level set, and draw a bunch of gradients as you move through each previous gradient infinitesimally. This is your desired path.
Huh? You're leaving out the most important part here. What do the gradients look like in the picture?
Just perp to tangents of the level sets.
20:53
OK, that's what I wanted.
I'm leaving it out because that's not my question.
The question is : however, is this going to be time-minimizing as well as steepness-maximizing? I mean, I want to minimize time + maximize steepness.
Is that going to be the same path?
Ah, your question is what I waas about to ask.
I don't think it is.
Does following the gradient even necessarily take you to the top of the mountain?
oh, yes. No.
20:54
Excellent follow-up question, which appears in my diff geo notes. When is the steepest ascent path a geodesic (shortest path)?
BTW, shortest in time is a bit vague ... are you moving at a constant speed perhaps?
Yeah. Sorry about that.
Having just climbed a lot today, I can tell you that I walk faster when it's shallower or downhill than when it's steep :D
I often ask my students to design a surface where following the gradient path takes an arbitrarily long distance to get you to the top.
Oh, I think I can visualize one.
Sure. That's very possible.
Take a spiral. Cross it with $[0, 1]$. And now make the ends of the spiral shallower than the center (into which it spirals inwards).
20:57
Right idea. Good :)
Next?
Let me see. I forgot what the next question was. I think I have written these down, wait a second.
"I have wrote"? Shudder.
Grammar-nitpicking while chatting is lame. :(
Mathematics should not be at the expense of correct language use. :)
I just answered your email, @Balarka.
I am looking.
21:05
Hmm.
Hmm? Goodnight, @MikeM.
Morning.
I'm being bad today.
No comment.
I don't want to edit...
21:07
@MikeM: It just came across my FB feed that the head of the Benghazi committee has been "caught using private email server." F***ing hypocrites, all of 'em.
@TedShifrin Replied.
Got it. Onward.
BTW, you can avoid integration by considering $h(t)=t^{-k}f(tx)$ as suggested in the problem in the first place. Then you just show the function is constant because the derivative is $0$.
Weird, I don't seem to have written down any question at all. I am pretty sure I had 3-4 this morning.
@Balarka: You're too young to be having dementia.
That's my excuse.
21:09
@TedShifrin ah, alright.
Hello @TedShifrin Could you take a look at math.stackexchange.com/questions/1493072/… I am confused about how we get only $B=Pb$ and not $B= \pm Pb$. Could you explain it to me?
@user159870: So let's think about it. Suppose we do a reflection for $P$. What does it do to $T$, $N$, and $B$?
OK, I don't seem to have any other question right at this moment. By the way, derivative of dot product of function just follows from writing them down componentwise, use the algebraic formula for dot product to write it as sum and product of function from $\Bbb R^n \to \Bbb R$, and use sum and product formulas, right?
I have not tried it out, but I think it works.
@Ted Do you have any multivariable notes or texts I can refer to?
Yes, @Balarka. That's fine.
@Julian: I have a book :)
21:14
Name?
This is the book with the 112 lectures on YouTube. Multivariable Mathematics: ... [see my profile]
By the way, @TedShifrin: bananas was complaining about PDE's he'd have to do in his functional analysis and diffgeo classes this semester. I linked him to the lecture video where you talk about applications of PDE. I am sure he will be motivated by the minimal surface equation :)
@TedShifrin They are the same and only have the other direction, or not?
gg to 210$
21:15
What's even better is the differential forms proof that you get least area with given boundary curve @Balarka.
I mentioned you saying it's a challenge problem (for next sem). Is that going to be in chapter 8?
@Ted I am think I am going to just stick to taking notes from you lectures
Book prices are obscene. I do not defend them.
21:17
@Julian You're planning on too many things at once. Algebra + Topology + Calculus + Category theory...
@Julian: It's the exercises that are most important. They aren't in the lecture. :(
Just pick one (or at most two) of the things you want to do.
@Balarka: I'm trying to slow him down.
Then $210 it is
21:17
You're very good at that, @TedShifrin. :P
Hang on, Julian. Work on Spivak first, and then we'll talk about my stuff.
Or is what I have written wrong? @TedShifrin
Sorry, @user159870: I'm a bit swamped here.
I must choose what I love @BalarkaSen that is why there is so much
@Ted Alright
I'd love to learn homotopy type theory right now. But I am not learning it.
21:18
@Balarka: I guarantee that when you get through the inverse function theorem stuff, you'll love transversality and differential topology, not even mentioning differential forms.
Yes, me too. And you're book is not exactly like a mechanical engineering-type mult. calc. text. I am enjoying what I am doing right now too.
Well, duh @Balarka :P
I want to get to the part where you define the Frenet frames, for example.
Well, to really play with that stuff, you need the diff geo notes (which are free). :P
Maybe after I finish calculus :)
21:21
So, @user159870, what you put in your post is right to a point. We have $T=Pt$ and $N=Pn$. Instead of using cross-product, what happens if we use $N'=-\tau B$?
Oh, @Balarka, you meant just that one exercise you asked about?
No, I mean, you define the Frenet frames later on the chapter on curves.
I wanted to mention Fary-Milnor, 'cuz it's so cool, and there are higher-dimensional versions that involve more interesting topology.
@Ted: You get what you pay for!
@TedShifrin There's an 11 y.o. diffgeo guy here too, as I recently discovered. We're corresponding, and I think he may tell me a rough sketch of the proof of Fary-Milnor.
Well, @MikeM, are you referring to the ridiculous price or the free price?
21:23
The ridiculous handwriting for free. ;)
Oh, with Crofton's formula the proof is super-cool (basically baby Morse theory, but max/min calculus stuff).
No, @MikeM, the diff geo notes I'm referring to are nicely TeXed. The undergraduate text.
That proof, too, is in my diff geo notes, @Balarka.
I glanced through a print copy of the book I can't afford and I was impressed. But that doesn't mean I can afford it.
Oh, ok.
@Balarka: I only define $T$ and $N$ in the text. The Frenet stuff is in the exercise.
21:25
@TedShifrin Yeah, I have noted. Well, I'll get to it today or tomorrow morning.
Oh, @user159870, I went too fast. Because torsion should actually change sign when you reflect.
Still trying to finish a frustrating SE problem instead of editing. One is less frustrating than the other.
And then I'm done differentiating. After finishing off the exercises I left out, going to skip chapter 4 mostly, but I think you also have something to say on smooth manifolds in the last section of chapter 4, right?
@TedShifrin This formula of $N'$ is one of the Frenet-Serret equations?
Yeah, @user159870, but I spoke too fast. We know curvature must stay the same, but we don't know whether torsion switches sign or not.
21:29
I am confused now... Always when we have an isometry the curvature stays the same? @TedShifrin
Yes, because curvature must always be nonnegative.
Ok. I understand. @TedShifrin
But I am agreeing with your result that the binormal is the negative of the original binormal when $P$ is a reflection.
So do we have to take cases for the determinant of $P$? @TedShifrin
Yes, it seems that $T=Pt$, $N=Pn$, and $B=(\det P)Pb$. If we do a reflection, a right-handed basis turns into a left-handed basis, so this has to happen.
I don't know a formula for $Pv\times Pw$, but I can figure it out by dotting with a third vector and then using the property of determinants.
Maybe I'm being stupid at the moment.
21:58
@TedShifrin do you know the meaning of an arrow that looks like $)\longrightarrow$ but without the gap at the base of an arrow. It isn't a hooked arrow and is without any doubt a bracket shape and size, used on a diagram in the context of defining quotient modules? If you do, please share.
22:42
Also if a book doesn't define a ring as additively commutative it still is just they've not bothered to mention it right?
23:26
@AlecTeal If you share an picture, we might be able to clarify.
@AlecTeal Rings with $1$ are automatically additively commutative. =)
Proof $(1+1)(a+b) = (1+1)a+(1+1)b=a+a+b+b$ is also equal to $(a+b)(1+1)=(a+b)1+(a+b)1=a+b+a+b$, cancelling $a$ to the left and $b$ to the right gives $a+b=b+a$. TA-DA!
23:46
Hi @TedShifrin
01:00 - 19:0019:00 - 00:00

« first day (1907 days earlier)      last day (3411 days later) »