« first day (2401 days earlier)      last day (2914 days later) » 

00:00
were given that the complement of each separate set of squares is connected
Just checking if the username is changed
hey @Semi
OK, that's perfect
It's Don't disturb
Yes, we can see.
00:01
actually I had to refresh my browser for it to update
So, please don't disturb me ever again
it showed up in the side bar on other channels though
@Then okay
@Then we won't
that was odd...
00:02
@Don'tdisturb You'll have to fill me in on what I did sorry
i want to do maths but im watching a dog pop balloons for a world record...
I'm a bit amused at a campus-wide email I just got.
@Akiva I propose the following solution
Apparently they're renovating one of the buildings, and for whatever reason the conditions are such that it's louder than they expected.
Leading to the following bolded line in the email: "There are no public safety concerns related to the venting. It's simply very loud."
what about tinnitus!
00:05
Somehow, the phrase "It's simply very loud" is a bit comic coming from an official university email.
Consider a "sharp loop" to be a sequence of $k$ squares $S_i$ such that $S_i$ is adjacent or diagonal to $S_{i+1}$ and $S_1 = S_k$
Notice that this is the only possible formation that divides the plane into 2 unaccessible parts
Also, notice that the sharp loop must all be the same color
However, if the two squares are divided by a sharp loop (the only case that the path doesn't exist), then it contradicts the fact that "There exists a path avoiding the blue or orange", because this sharp loop is only one color
therefore, these two squares must not be divided by a sharp loop, and thus there must exist a path between them
Q.E.D.?
An interesting contrast: Playing a LoZ game in one window while having the soundtrack to the most recent Doom going on in another.
which one?
Link's Awakening
oh, haven't played that one
have you played OoT?
00:12
Sure.
I had that pretty much right as it came out.
Link's Awakening was the first one I ever played, though.
uhh
not sure you would care at all
So definitely a major nostalgia factor.
but I actually wrote a hack for that game in MIPS assembly
00:13
Nice.
@ZachHauk You could also have an infinitely long line rather than a loop
it randomizes the get item table by generating a bijection based on the filename
Madness.
But I suppose that works. I'm just slightly worried about whether or not we should prove that these are the only structures that divide the plane into disconnected pieces.
in fact, the file name is too long, so I have to hash it to a smaller number using the in game RNG
00:14
@ZachHauk (Well, we're not given that the complements are connected, only that X and Y are in the same connected component)
Best part of the present setup: The Doom music will be raging along, then get to a quiet bit...and oh hey, there's the (much quieter) LoZ music again.
The contrast amuses me.
i've never played Doom
my friend loaded Doom on his calculator, though
a port of it at least
I've played the original Doom via Dosbox.
Didn't really have a chance to play it when I was younger.
00:16
you can download it there if you're interested. and hey, someone commented
I haven't played the newest one, though.
00:41
what's all this music on calculator ??
we are able to play music on calculator @ZachHauk @Semiclassical
?
if you have a sound chip then i guess
ohh..model of calculator
hey @Ted
Akiva found that infinitely intersecting problem thing pretty quickly
00:47
@MickLH do me a favour please. The next time you'll see stuff being told here around just ask them what they created in mathemtics, what their original ideas are.
Unique ideas one cannot find elsewhere. Only one single original idea that revolutionized any area of mathematics.
he got the same solution except cosine instead of sine
@Don'tdisturb just because someone hasn't done research yet doesn't mean they're inferior...
You can do it more geometrically with closed curves, even :)
I was reading an article , it involves poisson brackets , but i am unaware of that ,where should i start reading about them to get intuition...
00:49
hi @Semiclassical
The route I know to Poisson brackets is analytical mechanics.
@ZachHauk Here is a different point: when you're dedicated to mathematics in the most profound sense, you don't have time for s*it. I'm too tired of seeing so much envy around.
@BAYMAX — typically some experience with symplectic manifolds and Lie bracket helps.
Start with Newtonian mechanics, then make the jump to Lagrangian mechanics, and finally to Hamiltonian mechanics.
At which point Poisson brackets show up naturally. That said, this is probably not the route if you're wanting to appreciate them from math directly.
00:51
@ZachHauk I don't do it for me now, I do it more for younger people that think like me and that can be kept unfortunately in a range of very low performance and wrong beliefs about what they actually can do
Intuition is hard to buy cheap!
Main reason I like Poisson brackets in that perspective is that it connects with the commutator bracket in quantum mechanics.
That's Lie bracket :)
Some users here should be banned for ever for spreading so much wron belief about how mathemtics should be approached.
rightly said @TedShifrin
yes @Semiclassical
00:52
eh, maybe so. but you'd never bother to say it in physics.
$\{x,p\}=1\rightarrow [\hat{x},\hat{p}]=i\hbar$.
Be humble, be humble (a profound hypocrite expectation), just as much as you cannot ever get their performance or ever dare to over pass it.
I believe Balarka is a pretty humble guy
He's grown up a lot ...
Anyone explain me what topological conjugate is?
As to why the bracket should be imaginary in quantum mechanics....oh look over there, a convenient distraction.
00:54
Topologically conjugate?
what are $x$ , $p$ i guess continuous functions in $C_{\infty}{M}$ , $M$ is a set @Semiclassical
@Ted Now that Balarka's all grown up you have me instead!
yes
@ZachHauk You're best as long as you come to them like a begger to share their knowledge with you. But you don't have to do that.
Typically, you'd take $x,p$ to be canonical coordinate and the (conjugate) canonical momentum
00:55
Basically, you can make one picture look like the other by a homeomorphic change of coordinates.
You have the power to perform better than all of them.
@Simple Context?
eh, Balarka was smarter than me at this age
Um, gee, thanks, @Zach.
Comparisons are rarely helpful.
00:55
@ZachHauk He might be 70 or 80.
Ted still has to deal with Balarka
@Don'tdisturb He's 17
And you, DogAteMy :D
I have learnt Lagrangian , Hamiltonian 2 years ago , i have to go through them now ... ha ha :) @Semiclassical
@Semiclassical this is a word on my ode text book
And me, come to think of it.
Oh, nice.
00:56
@ZachHauk that's my whole point. You don't have to be like anybody, you must be like you, the version of you passed through a lot of work.
I can summarize it pretty well, then.
It's weird having some people on ignore ...
In Hamiltonian mechanics, you've got the Hamiltonian function $H$ (which in a lot of cases is just $H=p^2/(2m)+V(x)$.)
@Don'tdisturb I want to star what you're saying but I can't decide which ones. Maybe all of them
The time evolution of the position and momentum is then determined by Hamilton's equations: $\frac{dx}{dt}=\frac{\partial H}{\partial p}$ and $\frac{dp}{dt}=-\frac{\partial H}{\partial x}$.
00:59
@TedShifrin Balarka will have been smarter than me at my age (given that he's smarter than me and younger than me), yes
(I should probably be doing q not x, but w/e)
ohh @Semiclassical $q$ , the generalised coordinates .. right or wrong??
@AkivaWeinberger OK
Find how much gallon of paint it takes to paint a room. If the lenght, height and width is given and also we know that it takes a gallon to paint 350 square feet of the room, why do we need height if width is the height?
01:00
Now I'm really gone.
I was not commenting on silly comparisons, DogAteMy. I was saying I still have to deal with you.
@MATHASKER 3D room, not 2D
@TedShifrin Ah, I see
Yes, you still have to deal with me :D
oh but still why do we do Lenght * height + lenght * height + width * height + width * height
is width in this case thikness?
Moreover, Hamilton's equations determine the time evolution of any function $f(x,p)$ through the chain rule: $$\frac{df}{dt}=\frac{\partial f}{\partial x}\frac{dx}{dt}+\frac{\partial f}{\partial p}\frac{dp}{dt}=\frac{\partial f}{\partial x}\frac{\partial H}{\partial p}-\frac{\partial f}{\partial p}\frac{\partial H}{\partial x}$$
No, no. You're painting the ceiling or no?
01:02
no I don't think we have to paint the ceiling in this case
@MATHASKER Each of those terms is one of the four walls. Each wall is a rectangle, right?
but if we had to would we do lenght * width 2 times?
yes
one time.
Opposite walls have the same size.
That's a rather interesting looking structure, and in fact that's just the Poisson bracket!
01:03
You never do length*width
ya i get why we do it two times
why
Ceiling, DogAteMy
I thought we weren't painting that
Namely, $\{f,g\}=\dfrac{\partial f}{\partial x}\dfrac{\partial g}{\partial p}-\dfrac{\partial f}{\partial p}\dfrac{\partial g}{\partial x}$.
Nor the floor
01:04
ya we don't have to paint floor or the ceiling
Right. There was an if we were in there ...
Just replaced $H$ by $g$ in the last equation?
@Semiclassical
Anyhow, I'm out.
One wall is length*height. The next wall is width*height. The next wall is opposite the first one, so it has the same size, so it's length*height again. The last wall is opposite the second one, so it's width*height.
The definition from the book I have is: suppose $X'=AX,X'=BX$ have flow $\Phi^A$ and $\Phi^B$. These two systems are topologically conjugate if there exists a homeomorphism $h:R^2\to R^2$ that satisfies $\Phi^B(t,h(X_0))=h(\Phi^A(t,X_0))$. The homermorphism $h$ is called a conjugacy
01:05
Either draw a picture, or look at the walls of the room you're in…
Sure. I'm writing it in generality, but in the above we indeed have $\frac{df}{dt}=\{f,H\}$.
nice @Semiclassical
what is homeomorphism?
Is there a way i can save selected messages so i can refer to them later or something like that ??
@Simple A continuous bijection (with a continuous inverse)
01:07
I know it from Wikipedia, not sure how it relate with ode
The continuous inverse part isn't necessary in this context, though, I don't think
why is it width * height, is width in this case the space betweeen the two lines of the wall?
@Simple Essentially you warp the plane in some way
I think you can bookmark it if you look at the transcript instead.
Alternatively, you can search the transcript to track down the message if you remember some of it.
it is a transformation?
@MATHASKER Look at the floor. The floor is a rectangle. One side (edge), the longer side, we'll the length. The other side we'll call the width. The distance between floor and ceiling is the height.
@Simple I suppose, but not necessarily a linear transformation
("Edge" as in, it's one of the twelve edges of the cube-shaped room you're in.
Do i have to star them but by doing that it will fill the righ side area in the chatroom @Semiclassical
Well, rectangular prism.)
ok, thanks
01:10
Scroll to the top of the page and click 'transcript' to see the messages posted so far today.
There's then an option on the right to 'bookmark a conversation'
What's neat about the above formulation is that it has a direct quantum-mechanical version, namely Ehrenfest's theorem: $\dfrac{d}{dt}\langle \hat{A}\rangle = \dfrac{1}{i\hbar}\langle [\hat{A},\hat{H}]\rangle$
How do you see what you've bookmarked? Is it on your chat profile?
ohhh
ok thanks
Not sure, come to think of it. @AkivaWeinberger
is there a way I can like write the questions correctly and send them here like if i want to square something like is there a way like in the main site where i can square instead of doing 2^2
01:12
See the Latex in chat link in the room description.
i did creae a bookmark but i htink all the messages from a certain point till the end were there even some funny conversations :) @Semiclassical
Ah, weird.
I have the start chatjax and render mathjax stuff bookmarked
I think you're supposed to be able to pick a section of messages, but I haven't used it in a while.
Not sure what you're after, then.
but when I write messeges like 2^2 will it so and 2 squared once I hit render mathjax
01:15
not unless you put $'s around it.
$2^2=4$ being an easy example
ohh ok
it's ok , it was very nice that you told about the Hamiltonian .. thank you @Semiclassical
Poisson brackets as I'm given them above are pretty limited in motivation, since the second argument is only ever $H$.
oh alright it works thanks
But when you've got more than one canonical coordinate, then you start being interested in brackets between different coordinates.
01:16
$\sqrt{\phantom{If you can read this, your LaTeX isn't on.}}$
ya @Semiclassical
In particular, recall the angular momentum $\vec{L}=\vec{x}\times \vec{p}$.
Yes , i do remember that !!!
$cos(x)(tan(x) + sin(x) cot(x) = sinx+cos^2(x)$ prove that these are equal
For this one I got upto
If $\vec{x},\vec{p}$ have canonically conjugate position and momentum components, then their Poisson brackets (appropriately defined for the case of 3 conjugate variables rather than just 1) are $\{x_j,x_k\}=0$, $\{p_j,p_k\}=0$, $\{x_j,p_k\}=\delta_{jk}$.
01:18
$cos(x) + sin(x) + sin(x)/cos(x) + cos(x)/sin(x) $
LaTeX tip: Write \cos rather than cos. It's the difference between $\cos x$ and $cos x$.
Similarly for several other functions.
So the brackets between the 6 variables in phase space vanish unless you take a pair of conjugate variables.
Im stuck now, can I find a common denominator and simplify $\sin(x)/\cos(x) + \cos(x)/sin(x)$
wait whats a conjugate
Classical mechanics stuff. Not relevant to what you're doing.
01:20
oh
cause I heard my teacher sayin conjuagate
"Conjugate" shows up in a lot of contexts, alas.
oh, does classical mechanics have to do with like electric engeenering stuff?
cause it sounds kind of like mechanical
Well, here it corresponds to (for instance) how momentum and position are related.
But in thermodynamics, you would say that pressure and volume are (thermodynamic) conjugates.
It wouldn't surprise me if that language shows up in EE as well.
Wow i think this is your favourite subject @Semiclassical
It's one I know well, yeah.
01:24
Yeah it shows up in nonlinear circuits @Semiclassical
Moreover: If you work through the (rather tedious) algebra, you find that the Poisson brackets between the angular momentum components are neat: $\{L_j,L_k\}=\epsilon_{jkl}L_l$.
(I may be off by an overall constant.)
Please don't mind , is there any text you can suggest , so that i can review them ... i know it sounds awkward but i think reading a lucid text is very much helpful in understanding ..
Then can i ping you regarding this @Semiclassical
The conjugacy between momentum and position, anyways, is a big deal in quantum mechanics: It's the origin of the Heisenberg uncertainty principle!
@ZachHauk Here's a quick exercise: find two disjoint closed subsets of $\Bbb R^2$ whose distance is $0$. (The distance between two sets is the infimum of the distance between a point of one set and a point of the other.)
are u ok @Semiclassical
01:28
oh sounds interesting I isnt quantum mechanics like physics or something...I didn't know math was that related with physics lol
I'm good, though I'm not sticking around much longer.
closed as in
complement is open?
@ZachHauk yes. or equivalently, closed under the operation of taking limits
(in a metric space)
ok @Semiclassical
a union of two closed sets is still a closed set, right?
01:30
any kind of text if you could suggest which relates Hamiltonian , and the Poisson brackets , it would be very much interesting ??@Semiclassical
The one I learned this stuff out of was Goldstein.
I don;t know why i like a lot of references ?
But a good analytical mechanics textbook should have it.
Classical mechanics by Goldstein ??@Semiclassical
@ZachHauk yes
01:31
That said, I'm not sure how far a standard analytical mechanics book will go. Some might not get as far as Poisson brackets.
Yeah.
Goldstein definitely does, though.
@AkivaWeinberger that's really nice
Nice i bought it absent mindedly some months ago .. now it will help... wow @Semiclassical
hah, that's handy
Yeah... can i ping you if get doubts regarding this @Semiclassical ?
Anyways, to connect this back to what you were interested in earlier: The above algebra is known as the $\mathfrak{so}(3)$ algebra. (It's also the $\mathfrak{su}(2)$ algebra. There's no difference between the two in this case, but with higher dimensions that changes.)
Sure.
What's important is that said structure is relevant in both quantum mechanics and in classical mechanics, though of course the meaning of stuff like position and momentum isn't the same.
01:33
$So(3)$ algebra , is it related to groups ?
@Akiva is it even possible?
@ZachHauk yes
(-:
I think some of that may show up in Goldstein as well.
To be sure, though, Goldstein is a physics textbook not a math one.
So you won't get a symplectic manifolds perspective.
01:35
I am new to manifolds as well.
I'm not really that versed in them either, tbh. Hence why I described things in terms of physics.
umm
i don't think this is right
Okay, out for now.
That will do i think , understanding them Physically
but maybe the sets $1 \cup \frac{1}{2} \cup \frac{1}{3} \cup \dots$
01:37
out for now ??
and $-1 \cup -\frac{1}{2} \cup -\frac{1}{3} \cup \dots
$
@Akiva
neither of those are closed
I mean, I'm leaving for the time being.
why aren't their complements open?
Yeah,good bye @Semiclassical
01:38
(also just say $\{-1,-1/2,-1/3,\cdots\}$ and $\{\cdots,1/3,1/2,1\}$)
oh, because it's an infinite union
@ZachHauk $0$ is in their complements, but there is no open disk around $0$ not containing any $\pm1/n$
For a system of differential equation $X'=AX$, when $A$ has eigenvalues $\lambda_{1,2}=a\pm ib$, the general solution is $X=c_1u(t)+c_2v(t)$ where $u(t),v(t)$ are two linearly independent solution. My question is where is $i$ since $e^{it}=\cos(t)+i\sin(t)$
@Simple if it's a $1\times1$ system, where would $i$ be?
01:41
What are the formal prequistes for Measure Theory
You can't really give formal prereqs per se
But a good standard is chapters 1-7 of Rudin
(Baby Rudin)
I think in the preface of Big Rudin it basically says you ought know that before starting it
@arctictern not sure, do you mean we have second order ode>
Daminary also I wanted to learn the Lesbigue Integral for futher exploring Fourier Analysis
@Simple disregard what I said, doesn't make sense
01:44
Yeah, for reasons I won't know until next quarter, you kinda want measure theory and Lebesgue integration before Fourier analysis
The book i'm reading is an intro-one
It for studies with beginning knowledge in mathematical analysis
What's it called?
Stein's lecutres on analysis volume 1 Fourier Analysis
Ah Stein is good
Goooooooood stuuuuuuff
01:45
@Simple If $u$ and $v$ are eigenvectors of $A$, write $x(t)=c(t)u+d(t)v$ for some scalar functions $c$ and $d$, then see what happens.
It's really good learning a lot
I think the first 3 volumes are supposed to kinda be independent of each other
Volume 3 loops back to volume 1
intersting
I've learned a lot of analysis form books 1 and 2
and shows you that the Fourier theory is deeply intertwined with understanding the Hilbert space $L^{2}(\mathbb{R}^{n})$
01:47
Intersting i'm not that far in yet
Lol today in office hours Schlag was all like "Hopefully in the last lecture or 2 of 209, Marianna will teach you about $L^p$ spaces, for which the ones we deal with are a special case by taking $\mathbb{N}$ with the counting measure"
I will have something like $\cos(t)+i\sin(t)$ in the solution @arctictern
@Daminark when I took 209 we definitely did $L^{p}$ spaces for longer than that
@Simple What happens when you plug $x(t)=c(t)u+d(t)v$ into $x'(t)=Ax(t)$ given that $Au=(a+bi)u$ and $Av=(a-bi)v$?
Eric aren't there reqiurments for understanding Hilbert space theory
01:49
@ZION linear algebra, analysis
I mean, measure theory in particular because the canonical examples will involve integral norms
arctic tern I heard there aren't any requirements besides LA and Real Analysis
and yeah ^ linear algebra is important for everything pretty much
Do you have any idea why we tend to do functional analysis before measure theory?
idk man chicago's math major makes no sense
01:51
I still have $i$
This apparently isn't even a phenomenon that started in the Soug/Schlag days, even Sally's book has chapter 4 on functional analysis and chapter 6 on integration
Isn't it supposed to be MT before Functional Analysis
Yeah, normally
Our school doesn't play by the rules I guess
I mean I guess you don't really need measure theory to learn about abstract Banach spaces and to work with the sequences spaces
But it's handy
01:52
@Simple what do you get?
and that gives you a load of theory to work with.
yeah I agree it is handy
I was told initially the foundations for Measure Theory were a deep understanding of Point Set Topology, The Riemann Integral, and etc
We're definitely feeling it this quarter, because our problems are basically crunching out stuff with $\ell^p$ spaces, and the only books anyone in our class has so far found that don't just sweep it under the $L^p$ umbrella are K&F and Kreyszig
I guess? Idk how much point set I really thought about while learning measure theory for the first time.
Measure Theory seems to me a "So you want to be a math grad student" threshold
01:55
Semiclassical I thought that was real analysis
Suppose $A=\begin{pmatrix}a&b\\c&d\end{pmatrix}$ with $\lambda=a+ib$ and $u=[i,1]^T$, then $Au=(a+ib)[i,1]^T=[ai-b,a+ib]^T$ @arctictern
Nah, I did real analysis as a senior.
That'd imply that $a+i b$ is an eigenvalue of $A$ and I know that's false unless you know more about $c,d$.
@Daminark idk anything about this in particular but apparently there was some conflict when trying to figure out foundations of QM about whether we should use an $\ell^{2}$ formulation or an $L^{2}$ formulation but then they ended up being the same lol
at least this is what a grad student told me last year
but that's why newer books just sweep $\ell^{2}$ under the $L^{2}$ umbrella apparently.
Eric but does Measure Theory have any applications to physics
I don't really know much about physics
01:59
Dang
Well Eric from learning Fourier Analysis what important tools and techniques did you pick up
@ZachHauk Want a hint?
Neither of the sets are bounded.
hmm, i have an idea maybe
Uhh I mean if you wanna do PDE stuff you need lots of fourier analysis
02:03
Yeah that's one of the things i'm aiming for
Fourier analysis is also a great primer for quantum mechanics, at least the wave equation side of things
@ZION like the fourier transform does nice things to differential operators
Interesting, interesting I heard it had applications to Number Theory
Which connects to what I said earlier, actually: In quantum, one likes to talk about a position representation and a momentum representation.
And the relation between them is just the Fourier transform.
ahh ok
02:05
@Akiva I want some kind of set that "asymptotes" the other
that is, they're never equal, but their limits at infinity are equal
Also if you want to understand how the laplacian works you're gonna need lots of fourier analysis
and understanding the laplacian is critical for a lot of things
@Ted do you have any psets for that linear algebra course?
@ZachHauk yes
yes what
to your comment that I replied to (hover over the gray arrow on my message, it will highlight your remark that I replied to)
02:16
as in, it's correct?
so I want the difference between those 2 values
to approach 0 over time
$1/x$ seems like a good candidate
but they cant share any values
I don't see how integers are relevant
so, state the two sets you have in mind
02:21
time to make up data for this lab
sorry i gotta do this dumb lab lol
please don't make up data for your lab
but
it's 8th grade physical science
and he doesn't grade based on data
he grades on the format of the lab
 
1 hour later…
03:24
This place is a ghost town now
I'm a Lone Ranger of projective geometry
Kneel before my tremendous projective duality

« first day (2401 days earlier)      last day (2914 days later) »