« first day (2664 days earlier)      last day (2358 days later) » 

4:00 AM
@Daminark Math textbooks and whiskey.
 
@user104729 I mean it seems like I found happiness in page 1 of "Hodge theory and complex algebraic geometry 1" so knock yourself out
 
It is pretty good book
I recommend reading it @Daminark
I would say if you have working knowledge in basic differential geomtery and basic complex analysis you can read it pretty good
 
I don't really know diffgeo all too well, it came up in the bootcamp but I found myself working more on probability and whatnot
I asked my complex analysis prof if we'd be doing Riemann surfaces, he said no but told me to look at Lang's Intro to Algebraic and Abelian Functions
 
Which looks aggressive but like something that I could actually handle after learning ring theory (maybe)
Hey @Forever
 
4:14 AM
Check claire's book
 
BTW, Demonark, when Karim says diff geo he doesn't mean diff geo. He means basics of manifolds and bundles and differential forms (so more like diff top for you).
 
it is very readable to me
@Daminark I am pretty sure you your able to read Claire's book with ease
yeh
I would say if you know bundles, differential forms, integration/orientations on manifolds, basic complex analysis you can read this book easily.
 
Ah, I see @Adeek and @Ted
 
hi @TedShifrin
 
Maybe I'll check it out then, complex analysis was good fun so complex manifolds sounds cool. Also hey Ted!
 
4:18 AM
Karim needs to stop calling things differential geometry when they aren't :) When you have a connection on a vector bundle, then call it differential geometry :P
 
okay haha
 
@Ted i too am bothered by this
 
@TedShifrin did you have to know a lot of Riemannian geometry for complex geometry btw ?
I am really enjoying analytic complex geometry. I find it more geometrically appealing than abstract algebraic geometry.
 
"birth is a curse and existence is a prison" - where is this from?
 
Some familiarity with the basics of Riemannian geometry will definitely help you when you're doing the Chern connection on a hermitian complex manifold or holomorphic vector bundle.
 
4:21 AM
it's from the good place
 
this seems very much in line with existentialism: "we are condemned to be free"
interesting
 
I see @TedShifrin
 
it's from a comedy tv show lol
 
Morning @TedShifrin
 
comedy is great for philosophy
see rick and morty
2
 
4:23 AM
That sounds like that philosopher that starts with a B
 
@Ted knowing me to whatever degree you do, do you think that'd be worth trying out?
 
@orbit-stabilizer thats show was one giant wtf
 
idt rick and morty wants you to take its internal philosophy seriously
 
Hi Faust
 
hows your teaching project going?
 
4:24 AM
I haven't seen it, but it looks like it talks about the absurdity of life
Finding meaning in a meaningless world
 
Demonark: First, I don't think UC teaches complex manifolds. Second, you should do a graduate course in basic manifolds first. I don't think it's a high priority for you, no.
 
@orbit-stabilizer its worth watching if you want a confused funny however many hours
 
Camus is big on the absurd
 
@Faust: Doing easy polar coordinates tomorrow, but I'm going to have them discover things about ellipses and why you get them when you slice a cone right.
 
4:25 AM
Yeah we don't, if I do it at all it'd just be self-reading from some book
 
Not so much absurd, @orbit-stabilizer. That's more Beckett and Genet. But Camus and Sartre were existentialists.
 
if you wanna read serious existentialism do that, if you wanna watch funny sci fi irreverence watch rick and morty
 
Eh, camus being an existentialist is debatable
 
sorry
Schopenhauer doesn't start with a B
 
I think one can learn something in every moment
 
4:26 AM
This reminds me of Monty Python. "I've come for an argument." "No, you haven't." "Yes, I have." ...
hi @PVAL
 
howdy @Ted
 
Maybe just Riemann surfaces then? Would that be more tractable? And also do you have a source you recommend? I know Balarka was doing Forster and someone once recommended I check out Miranda
 
Momento is a great movie undoubtedly, but the philosophy really makes it stand out for me.
 
@TedShifrin i had a question from a student of mine regarding maximizing the area of a triangle of two given sides which were not the base by choosing the angle between those two sides
 
Hey @PVAL!
 
4:27 AM
@te
@TedShifrin hah
 
They're also a modern book of Donaldson.
 
i realize it is $\pi /2 $ but im not really sure how to explain why
any insight?
 
Riemann surfaces is a better idea, Demonark. I don't know Miranda's book. Forster has a lot of serious analysis. You might check out Griffiths's beautiful little book for a less overwhelming treatment (no sheaf cohomology, for example), Introduction to Algebraic Curves ... more classical.
 
I think for a topologist, Donaldson is really nice.
 
i forgot i was supposed to read that little chern book at some point
 
4:29 AM
@Faust: What do you mean not the base? Any side can be the base.
 
It cover the same material as these classical books from a slightly different perspective.
 
sorry base is supposed to say hypotenuse
 
No, that's not right, either, because that only makes sense in a right triangle.
 
you the longest side of a triangle w.e its called
 
Oh @PVAL I didn't realize that was directed to me :P
 
4:30 AM
i think the question just said given to sides of a triangle
 
I think this is all irrelevant, @Faust. All you know is two sides $a$ and $b$. Let $\theta$ be the included angle. Then the area is $\frac12 ab\sin\theta$.
 
what angle between them to maximize the area
 
That's obviously maximized for $\theta=\pi/2$. It's also clear from geometry, because that's making all of $b$ the height (over base $a$).
 
why is it that value?
 
Consider $\sin$ on $[0,\pi]$. What's its maximum value?
 
4:31 AM
no i know the answer
where did you get 1/2 ab sin
how do i derive that
 
1/2 base height
 
Draw a picture. Make $a$ the base and drop an altitude. What is its length?
 
find height
 
oh im an idiot i got it
thanks
i just drew a couple of triangles and decide it had to be $\pi /2 $
 
@Daminark Here is the following list of books that I recommend
 
4:35 AM
Is Ahlfors a good book?
I liked his complex analysis book.
 
It's a bit dated, but I like Ahlfors better than most modern alternatives. I don't know them all. But I tried teaching out of Stein/Stakarchi and was highly disappointed.
 
@Daminark
laus Fritzsche and Hans Grauert From Holomorphic functions to Complex Manifolds.


Complex and analytic differential geometry.

Daniel Huybrechts-Complex geometry an Introduction

Hodge theory and complex Algebraic geometry by Claire Violin.
 
I mean the Riemann Surface book he wrote
 
Voisin
Oh, it's not a general R.S. book. It's highly specialized.
 
For complex analysis I recommend the following books

Carlos Bernstein Roger Gay Complex analysis.
 
4:36 AM
Unless I'm thinking of the wrong book.
 
I like learning out of the Stack Project book :P
 
how hard are pdes ?
 
Carlos book introduces operators in the right way
 
Depends whether it's an undergrad/engineering style plug-and-chug course or a graduate course using a year of grad real analysis.
 
If you want basic classical book in complex analysis I would recommend Freitag complex analysis he also has the solutions with it.
 
4:37 AM
@Faust: if you want a nice book that gives a good introduction, look at Fritz John's Intro to PDE book.
 
im supposed to take a class on pdes next semester but its be 4 years since i did diffrential equatons and im pretty worried
 
Fritz john is good
 
What book, Faust?
 
@TedShifrin I should take PDE at some point in my life.
I never took it before
 
Yes, complex geometry is full of PDE.
Hodge theory is PDE, for example.
 
4:38 AM
I see
cool
 
Applied partial differential equations
 
Also I would like to go in mathematical physics/complex geometry I heard Einstein equations are also PDE
 
by Richard Haberman
 
Everything is PDE
 
Even classifying topoi and motivic cohomology are PDE.
 
4:39 AM
Oh, I know that book. It's sort of exhausting. It's pretty computational, but there are bunches of different chapters doing the same thing over and over with slight changes.
 
@user104729 what ??!
 
i dont understand it at all
 
Motivic cohomology are PDE?
 
I'm joking :P.
 
i think our undergrad class here is out of evans
 
4:40 AM
You need to know basic separation of variables from ODE for sure, Faust, but you also need to know multivariable calc (Stokes's and divergence theorem).
 
hmm i learned all those things 4 years ago
 
Undergrad is out of evans?? wtf
 
Doesn't Evans require a solid knowledge of $L^1$ or $L^2$, Eric?
 
do you not need functional analysis for evans?
 
yes
 
4:41 AM
and i have a really heavy course load next semester with more analysis and topology so maybe its a bad time
 
I continue to say UC is f***ed.
Too much, @Faust.
 
jeez
I mean, we have functional analysis here as well, but it's a fourth year course
no time for pde's as well
*pde's out of Evans
 
@Ted i think we need the analysis sequence as prereq to take it
i have no idea cause i just skipped all that
 
I taught a year-long applied math course back in 1986-87. I based it on Strang's (then) new applied math book, adding a lot more exercises and proofs. But it was a great course, and I learned a ton teaching it. We talked about complex analysis, Fourier stuff, discrete Fourier stuff, PDEs and fundamental solutions, lots of cool stuff.
Stationary phase is so awesome.
I also loved thinking of wave velocity and the Kelvin angle for waves in water.
 
in 1D it's nice.
multidimensional applications of stationary phase are scary
 
4:44 AM
Yeah, I just did 1D, Semiclassic.
 
In mathematics, the stationary phase approximation is a basic principle of asymptotic analysis, applying to oscillatory integrals I ( k ) = ∫ g ( x ) e i k f ( x ) d x {\displaystyle I(k)=\int g(x)e^{ikf(x)}\,dx} taken over n-dimensional space ℝn where i is the imaginary unit. Here f and g are real-valued smooth functions. The role of g...
 
is that the thing for integrals
 
I also found it fascinating to understand and explain the Huygens phenomenon ... that the wave equation is so different in even and odd dimensions.
There's so much cool stuff that is "beneath" the pure mathematicians in this room.
 
"pure"
""mathematician""
 
4:45 AM
'mathematican"
darnit
 
LOL, Eric. Been driving on those narrow mountain roads? :P
 
I like applied things
 
I wasn't referring to you, of course, Eric.
 
since we mentioned to stationary phase, here's an integral
 
For me replace "beneath" with "over my head".
 
4:47 AM
$\displaystyle F(z,\lambda)=\int_{-\infty}^\infty e^{\lambda (zx-(x-1)e^x)}\,dx$
 
intuitive explanation for why taking the quotient of a non-normal subgroup isn't a group?
 
I'm not talking about stuff at research levels, PVAL, just undergrad levels.
 
Ok, probally a good call. i have to take an advanced linear algerbra class and topology and and a second class in real analysis i need to take one more course. my options are a 4th year number theory class a 4th year advanced odes class, an abstract algerba class on galois theory an applied abstract algebra class or a combinatorics class. any suggestions?
 
@orbit-stabilizer: Work out what you need for $(aH)(bH)$ to be well-defined.
 
for $z>0$, that integral is convergent
 
4:47 AM
Three is already too many, @Faust. Two should be plenty.
 
@TedShifrin thanks will do
 
and if I consider its behavior as $\lambda\to\infty$, I can do a stationary phase approximation
 
@Faust i took a graduate elementary number theory course. Not a fan.
 
How is this stationary phase without an $i$ in there somewhere, Semiclassic?
 
4:48 AM
Oh, is $z$ restricted in real part or something?
Still looks suspicious.
 
yeah i wish i could just take 3 but i have to graduate within a reasonable time frame so i cant take less then 4 classes :(
 
I suspect I'm using 'stationary phase' when I really mean 'steepest descent'
 
@Faust personally, i'd go with the galois theory and combinatorics
 
@Faust: You cannot do 4 maths. You just can't. You'll either fail a few or drop a few.
 
yeah, for now I just mean $z$ real and positive
 
4:50 AM
I kind of did that this semester haha
I ended up withdrawing from one
 
Is $\lambda$ pure imaginary, @Semiclassic?
 
last time i tried more than 3 math classes i actually died
 
"literally:
 
@orbit-stabilizer: I had a few super-strong advisees at UGA who managed 3, but most of my advisees found 2 medium-difficulty classes plenty to handle.
 
4:50 AM
"literally"
 
@orbit-stabilizer thanks for the suggestion
 
but like I said, I should probably have said 'steepest descent'
In mathematics, the method of steepest descent or stationary-phase method or saddle-point method is an extension of Laplace's method for approximating an integral, where one deforms a contour integral in the complex plane to pass near a stationary point (saddle point), in roughly the direction of steepest descent or stationary phase. The saddle-point approximation is used with integrals in the complex plane, whereas Laplace’s method is used with real integrals. The integral to be estimated is often of the form ∫ C ...
 
Oh, geez, Semiclassic. rolls a few eyes
 
@TedShifrin i did five this semester and it almost killed me
 
hey, it uses "stationary-phase method" in the first sentence
 
4:51 AM
@Faust goodness
 
and plus I eventually would want to consider the integral under analytic continuation of $z$ off the real line
 
@Faust: You just have to be realistic. Take more time and actually learn something and pass. You're not a super-star student. Take some non-math along with 2 (max 3) math.
 
Did you even have time to digest the information?
 
i think 3 hard classes is a spot where i can do it but im having less fun
 
To me, Semiclassic, stationary phase requires oscillatory ("phase").
@Faust: The courses are already getting harder, not easier.
 
4:53 AM
maybe I should really be saying Laplace's method
meh
 
i end up on the steepest descent page every couple months and think "huh cool"
and then leave
 
There's a beautiful year-long intro to applied math book from MIT by Bender and Orszag. I highly recommend it to @Semiclassic and @EricSilva.
 
the names tend to blend together in my head
yeah, I've seen it
 
Well i do have an ideaic memory and im reasonably smart so that helps with digesting the information, i currently have above 90% on all my midterms and close to 100% on all my assignments so im not doing badly its just topology and PDE's scare me
 
it's the standard whenever one talks about asymptotic methods in physics
 
4:54 AM
Anyway ill think about only doing 3 classes
 
There's plenty of stuff in there I certainly don't know, but I taught some of the stuff in my applied math course. (I think I stole most of my stationary phase stuff from there.)
@Faust: I'm not trying to be mean, but after 40 years of advising students, I think I know you well enough to make reasonable suggestions.
 
sounds cool
man i wanna learn something about physics
 
stationary phase approximation to me basically means "deform the integral in the complex plane so that the integrand looks gaussian"
and when it already looks gaussian then just go directly to the second step
 
I'm thinking of doing measure theory, measure-theoretic probability, mathematical statistics (all 3 are cross-listed grad courses), and a fourth year stats course for my first semester in fourth year. Does that sound reasonable?
 
Taking 4 classes can be possible, if one of more of those classes is "beneath" you (i.e you know the material already or at least are way more experienced/skilled than the majority of the class).
 
4:56 AM
Well, I think of it as Taylor series expansion around a critical point inside the exponential.
 
I don't know if taking those classes is such a good idea though.
 
No, @orbit-stabilizer.
 
oh
only 2 math courses
 
@PVAL-inactive yeah my analysis course has been really easy this semester only reason i survived
 
and 2 statistics
 
4:57 AM
Too much of the same stuff, and probably at least one should be a prereq for another.
 
2+2=4
2
 
I don't think they're the same things?
I"m a math/stats major
 
It's all the same flavor stuff, which I don't like for an undergrad — probability/stat is totally overlapping a lot. And measure-theoretic probability should have measure theory as a prereq.
 
I had a lot of undergrad credits when I started college and routinely only took 12 quarter hours.
 
in measure theoretic probability youd see a lot of stuff that's just rephrasing measure theory
 
4:58 AM
I think most everyone else had to take 16.
and most of that was in your major.
 
Well, I took enough credits as an undergrad to get a triple major. But I didn't.
 
@TedShifrin it just financially its really hard for me to only take 3 classes i have alot fo mortgages and you never know when a big expense may come up and im not too sure how long the money i have will last
 
But I don't advise students to be like I was ... plus mine was never more than 2-3 maths. I was doing all sorts of languages and humanities.
 
I think when I was taking 16-17 hours 5 of them were guitar performance.
 
@Faust: I get it, but 4-5 maths is just not sane and very risky.
 
4:59 AM
so that really wasn't a full course load even then.
 
yeah and i cant do language classes my broken brain makes them very hard
 
You risk failing or having a mental breakdown.
 
@Ted didnt u double major tho
 
I'm not telling you to do languages. But find something else for electives.
 
I don't know if 2 math 1 physics 1 CS is so much easier than 4 math.
 
5:00 AM
Eric, yes, French was the second, but I didn't get the degree because I didn't write a thesis. But I took 8 or 9 upper-division courses.
 
i guess i could maybe do a physics class thats below me
 
@PVAL: I wasn't suggesting throwing in hard CS/phys as a reasonable alternative.
 
ah see i have enough for a medieval studies major but im not gonna do a thesis
 
But it's at least of a different flavor.
 
Well not everyone is so cultured
 
5:01 AM
Right, we discussed that, Eric.
Undergrad is not meant to be grad school where you do nothing but math.
 
if there's a course you know you'll eventually have to take which will be pretty mindless, this might be a good time to throw it in
 
@PVAL-inactive this in fact sounds way harder to me than 4 math
 
@TedShifrin but i love math >.>
 
I always told my advisees to save some of their core (freshman/soph) requirements for junior/senior year when they had all their hardest math.
 
I think even that
doesn't sound as hard to me as having to do language or write philosophy
 
5:02 AM
@Faust it's hard to keep loving anything if you have to do it all the time
 
@Faust: If your courses get more sophisticated, you will not succeed. I don't know how much longer you'll love it if you start failing and go nuts.
 
lol
that would defintly suck
 
I took an upper division class on Nietzsche and almost instantly felt like the stupidest person in the class.
 
I'm done giving advice. Obviously the guys doing Ph.D.'s in math have abnormally great talent in math. Not all undergraduate math majors have that much talent.
 
5:03 AM
@Semiclassical im Autistic i could play a video game from when i was a kid over and over again for months and not get bored.
 
I took one on metaphysics though and felt that the philosophy people were in general way too weak at formal logic.
 
@PVAL i kinda like that feeling and i think its part of what i like abt taking humanities
 
and made some pretty inane arguments.
 
Well, I find philosophy incredibly difficult, @PVAL, so I would tend to agree. But a sociology or an econ or an intro biology or a survey of literature or intro bio. I dunno.
 
I took a game theory course which wasn't rigorous.
It was boring as hell.
 
5:04 AM
anyway thanks everyone gave me something to think about
 
Anyhow, I'm out for now. I've said my piece (but not my peace).
 
@TedShifrin thanks for listening to me and the advice
 
If you want to see incorrect proofs of a very standard piece of mathematics.
Find game theory notes
and search "Brouwer Fixed point theorem"
 
oh man i wanna see a class on incorrect proofs
 
"unproofs throughout the ages"
 
5:06 AM
you could play it straight as a math history/phil class
 
Those are interesting topics, but undergrad courses on them will make you want to break things.
or at least made me want to break things.
 
a class on breaking things
 
i'll admit, I probably couldn't take a philosophy course
i like particular philosophy sources
But i'm not really interested in philosophy as a wider subject
 
I did id
I constructed a crazy space
 
5:43 AM
OH MAN OH MAN OH MAN
@Daminark I had to come to tell you this
still here ?
 
Nagel's papers are great
here's one on moral luck
 
@PVAL-inactive here ?
 
kinda
do you want something of me?
 
I have to tell someone about this.
@PVAL-inactive It is really cool you can prove Cauchy integral formula using general version of Stoke's theorem and
closed form
 
Hey, is Durrett's book on probability good?
 
5:51 AM
That's the only way I know how to prove it.
 
really ? Not the way your taught in understand ?
 
or I would deduce Greens from Stokes.
 
undergrad *
 
The way I've always seen is via Greens or Stokes
 
The way I was taught in undergrad is you draw like figures with small circle etc
 
5:52 AM
using you prove the Cauchy integral theorem is zero using stokes or greens
 
it wasn't talking about forms etc
I see
Yeah definitely this way makes the most intuitive sense
 
I've seen it done using triangles because it's more general
 
6:09 AM
@Semiclassical Every day man on the block.
 
Would anyone happen to have an answer to this question: stats.stackexchange.com/questions/311693/…
 
6:44 AM
@Daminark you know what is a closed form ?
a form $\alpha$ is closed if $d(\alpha) = 0$.
If f is analytic one can show easily that $d(fdz) = 0$.
Stoke's theorem state the following
$\int_{M} \phi^* d\alpha = \int_{\partial M} \phi^* \alpha$
 
Well, that's not just Stokes'
That's Stokes' + the fact that exterior derivative commutes with pullback
 
so you get Cauchy integral theorem right away
yeah
It is trivial calculation though
To get Cauchy integral formula you have to do some little bit more work
Yeah this way of doing things is very good
I think complex analysis parts of it should be taught by delivering how forms work
and how those things behave
 
Here's the issue with doing that
Stokes' theorem requires your form to be at least $C^1$
And the proof that holomorphic functions are analytic uses the integral theorems
 
yeah
no
it doesn't use it
oh I guess yeah for power series
 
So yeah that's why it's not perfectly general unless you separately prove that holomorphic functions are C^1
 
6:52 AM
oh yeah
 
[Random]
We have infinite sets, we have "quasiinfinite" set like the infinite dedekind finite sets and we have finite sets
But what about the notion of a "quasifinite" set. That is, a finite set that actually behave like an infinite set
 

« first day (2664 days earlier)      last day (2358 days later) »