« first day (2749 days earlier)      last day (2273 days later) » 

10:10 PM
@MatheinBoulomenos Hi, hast du einen Moment Zeit?
 
@philmcole Hi, ja habe ich
 
Ich versuch grad die grundlegende Idee hinter Vektoren zu verstehen. Das hier chat.stackexchange.com/transcript/message/42822019#42822019 verstehe ich einfach nicht.
 
Hmm, ich verstehe die Frage nicht genau. Aber ich würde sagen, ein Vektor ist kein Tupel von Zahlen, sondern ein Vektor ist ein Objekt von einer Menge, auf der ich Operationen mit bestimmten Eigenschaften definiert habe und es stellt sich heraus, das man Vektoren (nicht eindeutig) durch ein Tupel von Zahlen darstellen kann
 
Ok. Also es gibt irgendwie so eine geometrische Definition von Vektoren, wo man sagt, Vektoren sind die Objekte, die sich in einer speziellen Weise unter einem Bezugssystemwechsel verhalten. Wird anscheinend auch oft die "physikalische Definition" genannt (z.B. hier en.wikipedia.org/wiki/… )
Nun macht das ja schon Sinn.
Aber wie weiß ich, ob ein gegebener Tupel auch ein Vektor ist?
Feynman sagt dazu "No, not every three numbers form a vector! In order for it to be a vector, not only must there be three numbers, but these must be associated with a coordinate system in such a way that if we turn the coordinate system, the three numbers “revolve” on each other, get “mixed up” in each other, by the precise laws we have already described." (wobei er mit Gesetzen die Transformationsgesetze der Rotation meint)
 
Ich bin kein Freund persönlich von so einer Definition, ehrlich gesagt (ich bin auch kein Physiker)
 
10:18 PM
Ja sie ist leider irgendwie schwammig.
Aber ich komme nicht drum rum, wenn ich Relativitätstheorie und später Quantenmechanik verstehen will, weil dort alles darauf aufbaut.
Eigentlich müsste es ja total simpel sein, ich meine Vektoren sind so fundamental überall inzwischen...
Trotzdem sitze ich davor schon seit mehrern Tagen und versuche irgendwie zu verstehen, wie man konkret prüfen kann, ob ein gewisser Tupel ein Vektor ist.
 
Also ich glaube man kann auch Physik verstehen, ohne sich die schwammige Physikermathe zu Herzen zu nehmen ;-)
 
Schon sicher das meiste, aber ich finde, dass ich die Grundlagen wirklich verstehen sollte (und Vektoren sind wirklich elementare Grundlagen).
 
Aber Feynman meint nicht, dass manche Tupel Vektoren sind und manche nicht. Es gibt keinen Sinn, zu sagen $(0,1)$ ist ein Vektoren aber $(-5,17)$ nicht. Eine Liste von Zahlen existiert einfach ohne Zusammenhang zu irgendwas. Um wirklich einen Vektor zu haben, brauche ich ein Bezugssystem, d.h. ich würde sagen, die Liste von Zahlen muss Element von einem Vektorraum sein (Physiker würden sagen, man muss das Transformationsverhalten kennen, aber das ergibt sich dann daraus.)
 
Es gibt z.B. den Tupel $(\cos(\theta),2)$.
wobei der Winkel von der x-Achse gemessen ist in 2D
 
Das Beispiel verstehe ich nicht
Sobald du das Standardbezugssystem fixierst, also $\Bbb R^n$ und einfach das Standarftransformationssverhalten festlegst, d.h. für eine invertierbare Matrix transformiere ich meinen Tupel von Zahlen durch Matrix-Vektor-Multiplikation, dann definiert auch jedes Tupel von Zahlen einen Vektor
Aber dieser Zusatzkontext ist eben wichtig
 
10:25 PM
Wenn man darauf die Rotation anwendet, transformiert sich das nicht richtig und sodas es kein Vektor ist. Ehrlich gesagt ich auch nicht ganz, habe das nur hier physics.stackexchange.com/a/241679/164677 gelesen
 
Ich würde sagen die verlinkte Antwort ist Quatsch. Eine lineare Transformation in $\Bbb R^2$ ist nicht eindeutig durch einen Winkel festgelegt
Nicht jede lineare Transformation ist eine Rotation ...
 
Ok...
 
Hello, can someone help me with my question ?
that is my question if someone can help me math.stackexchange.com/questions/2647338/…
 
Vielleicht hilft es ja, wenn ich noch den genauen Kontext nenne, wieso sich überhaupt die Frage ergeben hat.
 
Hello everyone!
 
10:31 PM
Hi @Dami
How is it going?
 
Hi @Daminark @Alessandro
 
Guten Abend @Mathei
 
hello @Daminark
 
Yo, I'm alright, how about you guys?
 
In der spez Relativitätstheorie definiert man den Vierer-Positionsvektor als $X=(ct,x,y,z)$, also einfach nur mit einer zusätzlichen Zeit-Komponente. Man will nun einen Geschwindigkeitsvektor definieren, kann aber nicht einfach die zeitliche Ableitung von $X$ nehmen, weil die Zeit abhängig vom Bezugssystem ist. Warum ist das relevant?
Wäre es nicht total ok, wenn jeder seine eigene Ableitung kriegt?
Die Lösung ist dann nach einer vom Bezugssystem unabhängigen Größe, der s.g. Eigenzeit, abzuleiten.
Ich verstehe halt nicht, warum das überhaupt nötig ist.
 
10:33 PM
The exams session is almost over and lectures will begin again next week, looking forward to it
 
OK, on my completely overly ambitious list of languages I want to learn, but probably never will because becoming a polyglot is hard (I assume), after Japanese and Arabic goes German.
 
@MatheinBoulomenos Perhaps
 
Arabic because I was in Nazareth yesterday and there were Arabic signs everywhere since that's an Arab city and I wanted to be able to read the Arabic
though I could probably just learn the letters
And, also, you know, lots of Arabic in Israel in general.
 
@0celo7 "the Dirac $\delta$ is a function that is $0$ everywhere except at $0$, but at $0$ it's $\infty$, but in such a way that $\int \delta = 1$."
 
@MatheinBoulomenos you have to try harder than that to hurt me
because it is a function, in a sense, and it is zero everywhere except the origin, in a sense
and that integral is 1, in a sense
 
10:39 PM
does this hurt you
 
it doesn't make sense at all unless you make precise what "in a sense" means
 
@Slereah that's an April Fool's joke
 
It makes a sense
 
That is a 100% official published science paper
 
in a sense
 
10:40 PM
@MatheinBoulomenos I am saying there is a sense in which it makes sense. This is a nonconstructive argument.
 
Also you can embed distributions in a function space :p
 
It's $\lim_{n\to\infty}\circ\ f_n$ where the $f_n$s get increasingly more bumpy and where, when you integrate it, the integral sign goes where the $\circ$ is instead of in front somehow
 
I know that distributions are a thing
 
you can embed distributions in the space of smooth functions on the hyperreal
 
Not really?
 
10:42 PM
nonstandard analysis is the solution to all things
 
It is
 
What's a good resource to learn some of the theory behind distributions assuming I already had a semester of functional analysis? We didn't cover them but both the functional analysis professor and the mathematical physics one kept talking about them
 
You just have ${\rm st}(\circ f_N)$ where $N$ is an infinite hyperreal and where the integral goes where the $\circ$ is
Nothing's changed except for some vocabulary
 
Well they are functions :p
 
@AlessandroCodenotti Grubb, Distributions and Operators
 
10:43 PM
Hell, even $\frac{f(x+dx)-f(x)}{dx}$ doesn't equal $f'$. You have to take the ${\rm st}$ of it first.
 
Thanks, I'll check it out! All this talking about distributions has made me curious
 
@AkivaWeinberger Hyperreal stuff is a bit confusing because there's like 3 different definitions of continuity
 
I'm only familiar with two? And the second one is weird
Unless you mean uniform continuity as one of them
 
I forgot all the definitions (they're in Robinson)
One is just the transfer of the usual definition of continuity
 
$$\nabla^2\left(\frac{1}{\|\vec{r}-\vec{r}\|}\right)=-4\pi \delta^3(\vec{r}-\vec{r}’)$$ is the big one when it comes to E&M
 
10:45 PM
@Semiclassical shoo
 
The other is the $st(f(x + \varepsilon) - f(x) ) = 0$ one
 
Green's functions are keeping me from glory right now
 
Lol, does that hurt you
Ah, Green’s functions
 
It hurts me because the rigorous theory of those things is awful and I have 3 appendices of my thesis devoted to it
3 unwritten appendices
 
10:46 PM
...help
 
It seems you've got appendicitis?
May I recommend performing an appendicectomy on your thesis
 
Maybe you should use tonsils instead
 
If anything I need more appendices
 
Chronic appendix deficiency
which I'm glad isn't an actual condition
 
Oh, in case anyone hasn’t seen it yet: golem.ph.utexas.edu/category/2018/02/mlab.html
That’s pretty good
 
10:50 PM
@philmcole Du solltest @0celo7 fragen, der ist hier der Experte zu Relativitätstheorie. But I don't think that the answer to the question "Why do we want to define qunatities that don't depend on your frame of refrence" depends on your precise definition of a vector
 
@Semiclassical Is it even a parody
I can't tell the difference from category theory
 
Refresh the page
 
@Semiclassical it makes as much sense as nlab every time I refresh!
 
is anyone even going to read this thing
I could write some BS and my advisor would probably sign off on it
 
10:51 PM
With nlab there are portions which I can get stuff out of
 
there's no way he's checking this convergence argument
 
just say that it's trivial
put it in a corrolary
 
"The following is left as an exercise for the motivated reader."
 
“The following is left as an exercise to whoever gives a s*** about it.”
 
think anyone will catch this?
 
10:55 PM
@MatheinBoulomenos I think it does, because another reason why it wouldn't work to take $\frac{d(X)}{dt}$ is because it wouldn't be a vector anymore. I don't yet understand why, but I read it here physics.stackexchange.com/a/265695/164677
 
"Hint: Throw a bunch of machinery at it"
 
It's a pitty that the answers are so short
 
@philmcole what exactly is your question?
@AkivaWeinberger Hint: See [12], [46], [47], and [69].
 
@0celo7 Basically why we can't just use the ordinary time derivative to get ourselfs the velocity four vector. Why is it relevant, that it is not observer dependent?
 
@philmcole I wondered this myself for a long time
The answer really is just that one gets something observer independent. This is new physics, so there will be some new definitions
 
10:58 PM
Great then I'm at least not alone :P
 
This is the "correct" definition of velocity in SR
 
I mean you can use any parametrization of the curve
It doesn't matter that much
But proper time helps
 
“The proof is left as an exercise to the writer when he can be arsed to do it.”
 
One of the answers in physics.stackexchange.com/questions/265594/… also mentions that then the velocity wouldn't be a four vector. I don't understand that either...
 
sounds like physicist mumbo jumbo
@Semiclassical translate
 
11:03 PM
@0celo7 Oh.. good :D
 
oh for the love of god
I need another appendix
;_;
 
@0celo7 move ur whole paper to appendices
 
I think that the answers to this question are pretty bad: physics.stackexchange.com/questions/241610/… one is implicitly claiming that any two-dimensional invertible linear transformation is a rotation and the other one is written by someone who thinks that mathematicians don't know what a dual space is or don't use them
 
@0celo7 [12] is a personal communication, [46] is in Hungarian, [47] is no longer published, and [69] is 470 pages and you didn't give us a page number.
 
@AkivaWeinberger [69] is Federer
it contains all math in some sense
 
11:06 PM
I don't know who Federer is
 
you're a pure soul
 
a garbage book
that's all u need to kno
 
@EricSilva someone should make "Hitler learns GMT"
 
It's a book made for people in grad school abroad that have to take an English exam but didn't study
 
lol
 
11:07 PM
@Daminark I don't get it
 
he's referencing a story that a prof told him
didnt federer basically write this book for one of his students
 
Another source, [98], is just a list of the axioms of ZFC
 
I want to say that the thing to do is start in a rest frame
 
almgren probably?
 
I should probably prove Theorem 2.3 already
 
11:10 PM
Where it clearly should make sense to differentiate (ct,x,y,z) with respect to time
 
Why do harmonic functions pop up in physics?
 
@0celo7 basically someone my prof knew back in grad school had an English exam, to translate some page of a book, and brought Federer
 
@orbit-stabilizer because physicists love Laplace’s equation
 
@orbit-stabilizer Long ago, the four functions lived in harmony. But everything changed when the physics nation attacked
(This is probably about when I should go to sleep)
 
@orbit-stabilizer physics is about waves and Laplace equation is a Wick rotated wave equation
 
11:11 PM
@0celo7 lol
 
@AkivaWeinberger wait where are you now? Isn't it 6 in NY?
 
Wick rotated?
 
@Daminark I'm in Israel on a choir trip
 
they happen to solve like the simplest elliptic minimization problem
 
Ah, makes sense
 
11:12 PM
Earlier today I sang in the Knesset, like two feet away from the head of the opposition party
who went to my high school, by the way
It was pretty great
 
Interesting. Thanks
 
@orbit-stabilizer I was kidding
@EricSilva doesn't know what he's talking about
 
Nice
 
In electrostatics, Gauss’s law and the definition of electric potential gives Poisson’s equation $\nabla^2\Phi = \rho/\epsilon_0$
 
I know basically 0 physics, so you could tell me anything
 
11:14 PM
@orbit-stabilizer but like Semi says, it's because of Gauss' law
 
Apparently Bibi went by in a car when we were about to leave but I missed it
(Netanyahu)
 
Where rho is the local charge density and epsilon0 is a number with units don’t worry about it
 
and because if you throw enough terms in the Einstein equations away, you get the Poisson equation
 
Hmm okay, I'll look into that. It just came up when doing Cauchy-Riemann equations
 
I've also been in the same room as Pence once but it was a really big room
 
11:15 PM
@0celo7 but what i said happens to be a way they come up
an important one considering it's the reason they frequently show up whenever you do something variational
 
Main thing is that, if you’ve got charges spatially separated, then rho=0 in the area in between
So you end up with a boundary value problem where solution has to be a harmonic function
It also comes up in flow of incompressible fluids, in steady-state heat flow, etc
 
I see.. I wish I knew more physics
 
@MatheinBoulomenos I'd wish I'd be able to distinguish between good and bad answers but I just don't have the background knowledge yet :/
 
The fluid flow example is a big one, since one case you could be interested in is airflow around a wing
In which case you might model some portion of that flow as being 2D
 
Hi,
 
11:20 PM
And then using conformal mappings from complex analysis is a big deal
 
@Semiclassical flashbacks
 
There is bus A which runs every 10 minutes and bus B which spend 15 minutes. What probability do I have while waiting for bus B, to see past 2 bus A?
 
It’s less pertinent in physics, where we tend to rely on separation of variables rather than conformal mappings
@0celo7 I wondered if that would trigger you :P
 
@Dattier 1.05
/s
 
I sat in on the undergrad E&M class today, and it was pretty much all separation of variables
 
11:22 PM
my geometric analysis lecture was pretty much all separation of variables tbh
 
And then went to the TA meeting for that class where we agreed that method of images problems are usually pretty boring
 
My PDE class was all seperation of variables :(
 
@orbit-stabilizer : a probability <1
 
@Dattier -0.05
 
11:24 PM
proba >0 @orbit-stabilizer
 
@Semiclassical isn't it just a glorified "guess and use uniqueness to check you were right"
 
oh damn son I wrote the proof already
lmao
 
If I've got some some submanifold $\iota : \Sigma \hookrightarrow M$ with some tubular neighbourhood with the appropriate normal bundle decomposition, can I write the metric tensor of $M$ as... $g = g|_{T\Sigma}+ g|_{N\Sigma}$
Or something similar
 
@EricSilva eh, a more principled view is that you’re using symmetry
 
@Datt $\epsilon$ where $\epsilon$ is a hyperreal number smaller than every positive real number and larger than 0.
 
11:28 PM
@orbit-stabilizer $0$?
 
what about $\epsilon^2$ ? @or
 
@Semiclassical idt this is really different beyond semantics, you can't make real guesses without symmetry to guide you
 
Eg if Phi(x,y,z) is the potential due to a charge distribution rho(x,y,z) then replacing z with -z gives another solution
 
@Slereah along the image of Sigma, yes
 
11:29 PM
@0celo7 Even in the whole tubular ngh?
 
@Dattier, I'm just joking.
 
@Slereah $T\Sigma$ isn't defined there
 
@or Ah
 
The tricky one is spherical inversion @EricSilva
 
@EricSilva my thesis contains old German GMT
 
11:31 PM
@0celo7 What about $T\Sigma_\varepsilon$, for projections of the normal bundle of coordinate $\varepsilon$
 
freshly translated
 
how much floor to a house without floor?
@or
 
I don’t really get how the Kelvin transformation for Laplace’s equation works
It evidently does
 
@Semiclassical the laplacian on R^n is conformally invariant
 
But it’s pretty weird
 
11:32 PM
@Dattier $K$, where $K$ is your favourite number.
 
Right
 
sorry, ill stop. I can't understand what you're saying @Dat
 
which is really the point i think
@0celo7 cool
 
@or ok
 
@EricSilva sure
 
11:33 PM
@orbit-stabilizer What if my favorite number is the number $K + 1$
 
Being able to do a problem using images is basically a trick of finding a combination which gives you the right symmetries
 
do you know there a big number in AP ?
the number of Goodstein
 
@EricSilva i remembered the formula saying it was conformally invariant on n/2-forms
am i using the wrong laplacian
 
im thinking laplacian on functions
on ricci flat guys(maybe?) that's conformally invariant
 
scalar flat
 
11:35 PM
ya that's it
some kinda flat
 
the number of Goodstein is a finite number K (you can show it in ZFC) but not in AP, so K+1=K in AP @Sl @or
 
laplacian on functions is conformal up to a scalar curvature term times some dimensional constant or something
 
En mathématiques, et plus précisément en logique mathématique, le théorème de Goodstein est un énoncé arithmétique portant sur des suites, dites suites de Goodstein. Les suites de Goodstein sont des suites d'entiers à la croissance initiale extrêmement rapide, et le théorème établit que (en dépit des apparences) toute suite de Goodstein se termine par 0. Le théorème de Goodstein n'est pas démontrable dans l'arithmétique de Peano du premier ordre, mais peut être démontré dans des théories plus fortes, comme la théorie des ensembles ZF (une démonstration simple utilise les ordinaux jusqu'à ...
 
@EricSilva $-\Delta+\frac{n-2}{4(n-1)}R$
 
there we go
 
11:37 PM
it seems weird to say "is conformally invariant" when you then restrict the curvature of the mfds? like you're actually only invariant under specific things
 
good work team
 
@EricSilva actually I don't think you mean invariant
 
the operators change but they change nice
 
yes, but that's not "the Laplacian is invariant under conformal transformation"
 
sure
 
11:40 PM
more like, if the manifold is scalar flat, and you make a conformal transformation to something scalar flat, then the new Laplacian is the old one with some changes :P
 
the changes are just like multiplying and dividing by a thing or smth
positive nonvanishing thing
 
powers of the conformal factor
 
@0celo7 was running into that earlier when trying to write out the 3D case
I say trying because I got bored of it and went on to something else
 
This business with harmonic conjugates, why not just pick a constant function? That'll be a harmonic conjugate
oh nvm
 
11:47 PM
@MikeMiller this is fair, i guess the terminology really is a little misleading
 
mfw the table of contents is over 70 pages
 

« first day (2749 days earlier)      last day (2273 days later) »