« first day (5259 days earlier)      last day (57 days later) » 
00:00 - 20:0020:00 - 23:00

00:00
"..a "compatible choice" of elements of $\sqrt{z^2 + m^2}$" by elements you mean single valued functions of this multivalued function?
yes, in the sense that I mean a single-valued function that at each point $z$ picks out one of the elements of $\sqrt{z^2 + m^2}$, in a "coherent" fashion
Ok thanks
Can you recommend me
any good books in complex analysis? Idk if it makes a difference
whether one studies math or physics
or if there is a special one for physicists
I've never opened a book on complex analysis, but when I took the respective course I found these online notes (?) pretty helpful phys.libretexts.org/Bookshelves/…
Bookmarked
thank you for your help
you're welcome :>
00:28
@BenSteffan I am back, because I have one more question xD
It's the root function
and now we are evaluating the integral
I will give you time to read
but once you are done, can we go step by step and discuss the text. There are things that confuse me
I'm done, but I'm not sure how much help I will be haha
why does the first exponential decreases ?
if $\vec p = a + ib$ the absolute value is a real positive number
and $|x|>0$
so you have a positive complext exponential, that can be written in cos and sin
or I am wrong?
there is something horrible going on with the notation in your source
00:38
what exactly ?
I don't understand why the write $|p|$, but evidently that's just supposed to be some complex-valued variable
you know why
otherwise how could you have "large imaginary values of $|p|$"
because as you can see $p$ is a vector right
since its in bold
shudders
well anyways, "large imaginary values of $|p|$" presumably means $|p| = iy$ for real, large $y > 0$.
then $e^{i |p| |x|} = e^{-y |x|}$ which is small
00:41
@BenSteffan we have $\vec p$ and by going to spherical polars we can write it the same way we write a vector from cartesian in spherical coordinates
Does it make sense now?
@BenSteffan Ahhh but you are right
I understand what it's doing, although I disagree with the particular choice of notation
if you see the the integral, the argument is $d|vec p|$
and the values range from minus infinity to infity
so if it can take negative values then it can certainly take values of the form $i\gamma$ or w/e
I'll just treat it like a complex-valued variable
And what if $|p|=a+ib$
?
I don't think that's relevant?
00:46
Ok
this part is analyzing the "vertical" parts of the curve I think
And then is the description for the other term
my understanding is
that if $|p|=i\alpha$ and $|p|^2 = -\alpha^2 >m^2$
then we have a negative root, so can be written as $i\sqrt{...}$
but I don't think that is the case
and I don't understand how it makes sense the notation $Im(\sqrt{|p|^2+m^2})$
@imbAF we chose a branch for the square root earlier, so this now makes sense
I understand the idea
I don't understand how the
two expressions come out
For what I know
you have a number for a given p value under the root
that's it
Maybe I am missing something
I don't understand what you're confused about
00:56
I really don't understand how we determine branches for root. For log I know that:

$log(z)= ln(r)+i(\theta + 2pin)$ where n=..-1,0,1,... which shows multivalued
so you picking $\theta$ being in 2pi
that is one branch
I understand this
I cannot describe something similar for the root
For the root you only get to choose between 2 values
can you give me an example other than \sqrt{-1}
You already have this on the real line: If $x > 0$, then it really has two square roots: $\sqrt{x}$ and $-\sqrt{x}$
in $\mathbb{R}$ this doesn't really matter: there's a canonical way to choose roots, namely just take the positive one
in $\mathbb{C}$ you can't do this anymore, since you cannot compare complex numbers
"...then it really has two square roots:" which has the 2 square roots?
$x$
say $x = 123.2381$ or so
or $x = 4$ :)
there are two real numbers that square to 4: 2 and $-2$
in the complex plane, there are two numbers that square to $-1$: $i$ and $-i$
there are two numbers that square to $i$: $1 / \sqrt{2} (1 + i)$ and $-1 / \sqrt{2} (1 + i)$
there's not much to it: if $x^2 = y$, then $(-x)^2 = y$ as well
01:02
I understand the explanation but in the exponential I am already having the root, meaning I am having the 2
or -2
yes, or -2
how are you distinguishing the 2? how do you know which one you have, which one to pick?
This is nasty stuff
it's one of the costs you have to pay if you want to harness the power of the complex world :)
Yeah but I am doing it on the fly
that's ok, you don't need to know that much of the gory details
01:05
In our example we IM(...)
you just need to be aware that there is such a thing, that it is a problem, and that it is solved via branch cuts
is it because we have the complex i in the exp?
we haven't gotten to the $\exp$
Ok but then why not have $\pm\sqrt{}$ but have $Im\sqrt{}$ ?
I think this is a valid question
ok, fine, I guess they're interested in the imaginary part because it's sitting in a $e^{-i \ldots}$
if you write $a + bi = \sqrt{|p|^2 + m^2}$, then $e^{-i t \sqrt{|p|^2 + m^2}} = e^{-i t (a + bi)} = e^{-ita} e^{tb}$ so the imaginary part is what determines asymptotic behavior, in a sense
01:14
And can one write $a + bi = \sqrt{|p|^2 + m^2}$ ?
Is that something possible
why would one not be able to?
it's just a complex number :)
cuz i thought that the result of a root is one of the two
ah
and you take 0 real part
you ignore the real part and just look at the imaginary part
for this argument
Ok I see
Then I will proceed from what we discussed
hopefully no more bumps on the road
thanks
welcome :)
 
1 hour later…
02:46
@Thorgott one thing I've noticed is that if $|X|<|\beta X|$ then $\beta X$ is not path-connected
In fact there exists $x\in\beta X\setminus X$ to which no sequence in $X$ converges
03:01
Well, I might want to assume that $X$ is realcompact
03:16
Actually I'm not sure how Moishe Kohan came to his conclusion without assuming that
03:56
The problem I have is from moving from path-connectedness to have a convergent sequence $x_n\in X$ to a point $x\in \beta X\setminus X$. I think this requires realcompactness
The realcompactness was the friends we made along the way
04:18
okay I see it now
 
3 hours later…
07:20
mow
hi
@Allie hi
yayyy someones here
whoah, woah. everybody can't just start talking at once
because im CONFUSED!!!!
btw i hope youre all having a nice evening
still morning but thanks, likewise
07:23
lol
time zones...
08:15
@Allie this got something to do with biotsavart law?
cuz i remember doing something like this back in the old days
theres this theorem u can perhaps make use of: $\vec \nabla\times (f \vec A)=f \vec \nabla \times \vec A-A \times \vec \nabla f$
 
3 hours later…
10:53
Is there a reason that the later couple is written as (p,X") and not as (X",p) like the formar couple (X',f)?
Is it to emphasize that the morphism f comes out of X', and p goes into X"?
 
3 hours later…
13:39
yeah, that's what I would think
in any case, it's a purely aesthetic choice
14:01
@Thorgott we were talking yesterday about when $\beta X$ is not path-connected
I came up with three theorems which give sufficient conditions for that
14:12
oh nice, tell me
I also found something yesterday that explained that $X$ is a path-component of $\beta X$ when it is, uh, paracompact, path-connected and of non-measurable cardinality (iirc)
I know what you are referring to, and one of the theorems I proved has this one as a corollary
If you know that paracompact + non-measurable cardinality implies realcompact, that is
0
A: Two questions on path connected spaces

JakobianAll spaces below are Tychonoff. Proposition. If $X$ is pseudonormal and $x\in \beta X\setminus X$, then there is no sequence $(x_n)\subseteq X$ which converges to $x$. Theorem 1. If $X$ is non-compact locally compact and pseudonormal, then $\beta X$ is not path-connected. Theorem 2. If $\aleph_0 <...

I've posted an answer under that same post
normality was replaced by weaker pseudonormality
theorem 2 is my idea about using cardinality to prove that $\beta X$ cannot be path-connected
answer by Joseph van Name is corollary of my theorem 3
theorem 1 is the one that Moshe Kohan was referring to, and what Eric Wofsey was commenting
proposition is a stronger version of what Martin Sleziak proved in this answer of his
in fact this reminds me that I can add more to this answer that I forgot about
 
1 hour later…
15:34
nice stuff
16:01
Hi
@SineoftheTime Can I show you something about physics for a moment which however concerns the extremes of integration?
why doesn't it integrate from $\pi/2$ to $3\pi/2$?
what's $\theta$?
16:08
wait ill update the graph
ok, why you should integrate between $\pi/2$ and $3\pi/2$?
I'm not like in this case?
take a look at $\theta$ in the picture
Are you sure it's the same?
when $\theta=0$?
16:29
im not understanding
Which image are you referring to?
@SineoftheTime at $\pi$
@Pizza it is not cosine(theta)
why?
Maybe it depends on how you take the angle
@Pizza the first one
@Pizza here $\theta$ does not vary as in the unit circle
So you integrate from $\pi/2$ to $-\pi/2$
otherwise you can integrate from $3\pi/2$ to $\pi/2$
or over any half period of the cosine, but you have to be careful about the sign
do you mean this $\theta$ here?
16:42
yes
or the other one, it's the same
@Pizza Why did you use cos(theta)?
take the one at the top
but the other $\theta$ is in another quadrant
and then it varies between $\pi/2$ and $3\pi/2$
or am i wrong
@Binky i writed dEx = dE cos theta
16:47
And why isn't it sin(theta)?
7 mins ago, by Sine of the Time
or over any half period of the cosine, but you have to be careful about the sign
@Binky
What
if you integrate from $\pi/2$ to $3\pi/2$ you get $\hat E=-2k \hat i$
@Pizza Can you show me the chart pls
16:50
@SineoftheTime yes
but the book also goes from $-\pi/2$ to $\pi/2$, and the solution does not come with $-$
so do I have to change the sign?
@SineoftheTime here
@Binky what is the chart
From the semicircle graph
That is, why are you taking the angle made with the x-axis and not the one with the y-axis?
Correct
Can you answer above pls
2 mins ago, by Binky
That is, why are you taking the angle made with the x-axis and not the one with the y-axis?
this?
Yes
16:59
@Pizza here, if you take $\theta $ "under the $x$ axis", do you see that if $\theta \in [\pi/2,3\pi /2]$ then $dE_x$ points towards the negative $x$ axis?
@SineoftheTime yes
so you have $-\hat i$ right?
If you used sin(theta) you had to integrate from 0 to π
@SineoftheTime yes
so $\hat E= -\int_{\pi/2}^{3\pi/2} dE_x \hat i$
which leads to $2k\hat i$
17:09
@SineoftheTime instead if I had been in this case it would have been positive:
$\theta=0° in \pi$
-> If I move half it will be -π/2 π/2
ah it was dEx, not just dE at the bottom sorry
@Pizza Clear?
17:11
@Binky I did not understand what you are doing
@Pizza here you get the same formula I wrote above, but without minus sign
which make sense since $dE_x$ points toward the negative x axis
maybe it's better if you ask to someone else in the physics chatroom
ok but if I wanted I could integrate from $-\pi/2$ to $\pi/2$ but with the $-$ in front of the integral
i mean in this case
3 mins ago, by Pizza
user image
$\int_{-\pi/2}^{\pi/2} \dots$?
yes
that would be fine
17:14
$ - \int_{-\pi/2}^{\pi/2} \dots$?
yes, you get $dE_x$ pointing toward the negative x axis
ah ok I think it's clear
\[
\theta = \pi + \frac{\pi}{2} = \frac{3\pi}{2}, \quad (\cos(\frac{3\pi}{2}), \sin(\frac{3\pi}{2})) = (0, -1).
\]

\[
\theta = \pi - \frac{\pi}{2} = \frac{\pi}{2}, \quad (\cos(\frac{\pi}{2}), \sin(\frac{\pi}{2})) = (0, 1).
\]

\[
\theta = \pi \pm \frac{\pi}{2} \implies
\begin{cases}
\theta = \frac{\pi}{2}, & (\cos(\frac{\pi}{2}), \sin(\frac{\pi}{2})) = (0, 1), \\
\theta = \frac{3\pi}{2}, & (\cos(\frac{3\pi}{2}), \sin(\frac{3\pi}{2})) = (0, -1).
\end{cases}
\]
@Pizza
maybe it's better if you ask a physicist to check if what I've said it ok @Pizza
tomorrow i can go in problem solving
anyway
17:16
@Binky that's not relevant to the discussion
Can you confirm that I was right to use $\cos(\theta)$? I don't understand what binky is saying
21 mins ago, by Pizza
user image
what is the current context?
$dE_x = dE \cos\theta$
@Pizza Assuming $d\vec E$ is correct, then $|dE_x|=|dE \cos \theta|$
If you take the angle with the y axis you have to use $sin(\theta)$
So it depends
17:20
@SineoftheTime ok
thanks !
np but I'm not that good at physics :D
So my explanation may not be clear
@SineoftheTime nor at maths
@Pizza How clear are mine?
I think you're confusing him
:(
Let $$Y = \left \{ f \text{ measureable in } \mathbb{R} : \int_{\mathbb{R}}|f(x)|^2 e^{-x^2}dx < \infty (\text{i.e. converges}) \right \}$$. Show that $f_n = \chi_{[n,n+1]} \longrightarrow 0 $ in $Y$
17:24
@Claudio do you have an idea on how to proceed?
yeah, I was writing but maybe Ill say it via words
I considered $\int_{n}^{n+1}e^{-x^2}dx$
looks goood
now I'd like to use the fact that e^{-x^2} decreases rapidly as n grows larger
but I can't seem to be able to bound the integral
note that $e^{-x^2}$ is decreasing over $[n,n+1]$, so can you esteeem $e^{-x^2}$ over $[n,n+1]$?
ohhhh wait I just realized something
17:27
@Binky did you mean this?
I've tried writing $\int_{n}^{n+1}e^{-n^2}dx$
but I thought I dont know how to intregrate this
but I just noticed the argument does no depend upon x ahahhaah
sillly me
happens to the best of us
@Pizza yes
@SineoftheTime straight facts my friend
mm okay so $dE_x = dE \cos(90 - \theta) = dE \sin(\theta)$
17:30
Yes
$\theta\in[0,\pi]$
mm
What is not clear?
what $\theta$ did you take
The one in the picture
I'm having trouble knowing how to see what $\theta$ varies between because I have $90 - \theta$
17:39
Its $\theta$
You have $d\theta$
Not $d(90-\theta)$
yes $dl = R d\theta$
Ok
So it confirms what I said
I didn't understand how you found the integration extremes
Ok
$$
\theta\left(\frac{\pi}{2}\right) = 0
$$
$$
\text{If I make a 180° turn, I arrive at } \pi
$$
$$
\int_0^\pi f(\theta) \, d\theta
$$
I didn't understand what changes between how we did it before and now, i.e. why are the extremes no longer $-\pi/2$ and $\pi/2$?
17:47
Ah
@SineoftheTime did you understand? (if you are following the chat)
I'm not following tbh
PDE experts any? or avid readers of Courant Hilbert vol II
@SineoftheTime I didn't understand what changes if instead of $\cos\theta$ I write $\sin\theta$ , in the extremes of integration
$$
\theta = 0 \text{ at } \pi
$$
$$
\text{Half path above: } \theta = -\frac{\pi}{2}
$$
$$
\text{Half path below: } \theta = \frac{\pi}{2}
$$
$$
\text{Counterclockwise direction: } \theta \in \left[ -\frac{\pi}{2}, \frac{\pi}{2} \right]
$$
17:51
ok, Isn't it like we did before?
This would be when you use $\cos(\theta)$
@Pizza yes
ok
up to this point I understood
$$
\theta = 0 \text{ at } \frac{\pi}{2}
$$
$$
\text{If I make half a turn counterclockwise, I arrive at } \pi
$$
$$
\text{The path is from } 0 \text{ to } \pi
$$
$$
\text{Counterclockwise direction: } \theta \in [0, \pi]
$$
Clear?
right
I understand, thanks
18:38
.
Does the adjoint representation "commute" with other representations? For example, is it true that $\text{Ad}_{\pi(g)} \pi_(X) = \pi_ (\text{Ad}_{g} X)$ where $\pi:G \to GL(V)$ is a Lie group representation and $\pi_*: \mathfrak{g} \to \mathfrak{gl}(V)$ is the induced Lie algebra representation.
@SillyGoose fix latex
I was trying to, but I am not sure waht the error is. this tex compiles perfectly normally in overleaf o.0
Does the adjoint representation "commute" with other representations? For example, is it true that $\text{Ad}_{\pi(g)} \pi_\star(X) = \pi_\star (\text{Ad}_{g} X)$ where $\pi:G \to GL(V)$ is a Lie group representation and $\pi_\star: \mathfrak{g} \to \mathfrak{gl}(V)$ is the induced Lie algebra representation.
okay i guess it doesn't like that i used asteriks
@SillyGoose Yes, because markdown happens first, and * is used for emphasis.
i see
im surprised i haven't encountered this issue before
@SillyGoose It is a problem if you have two *s in your comment. If there is not a matched pair, it will likely be fine.
Consider $*$, vs...
...first one $*$, and then another $*$. (This should break.)
Oh, or maybe $* \text{and then another} *$.
Huh... maybe that isn't the issue.
18:50
$\text{test}_*$ and $\text{test2}_*$
huh. $\pi_*(\text{Ad}_gX)$ and $*$
Also, you probably want \ast, rather than \star.
it seems maybe using "\*" could work around the issue as well
oh yes that also probably would work equivalently
\star is more elegant
@SineoftheTime Disagree. It is not the standard notation.
I did not say it's standard :D
19:12
@SineoftheTime do you know Agadmator?
@SoumikMukherjee I've heard about him
The dude who does recaps right?
Yes
I just played a game against him
no way
stalemate :(
Hehe, I was playing unnecessarily fast, I had plenty of time (6 secs)
I thought he was like an IM/GM
I've seen in fide he has $\sim$1950
19:16
I think he is a CM
In Italy CM= over 2000
in other countries is over 2200
Back in 2018, his videos are what that got me into chess. He used to do a chess series named what sorcery is this or something like that. I named my account based on that word sorcery.
Hi
19:32
$$
\sum_{n=1}^\infty \ln(27n^3 - 2n) - 3 \ln(3n).
$$
I need help pls
@Binky Okay... what have you done?
For example, have you thought about using properties of logarithms to simplify the summands?
$\ln\bigl(27n^3 - 2n\bigr) \;=\; \ln\bigl[n\,(27n^2 - 2)\bigr]
\;=\; \ln(n) \;+\; \ln\bigl(27n^2 - 2\bigr)$
$3\,\ln(3n) \;=\; \ln\bigl((3n)^3\bigr) \;=\; \ln\bigl(27\,n^3\bigr).$
$\ln(27n^3 - 2n)\;-\;3\,\ln(3n)
\;=\;\bigl[\ln(n)+\ln(27n^2 - 2)\bigr]
\;-\;\ln(27n^3). $
$\ln(n)\;+\;\ln(27n^2 - 2)\;-\;\ln(27n^3)
\;=\;\ln\!\Bigl[\tfrac{n\,(27n^2 - 2)}{27n^3}\Bigr]
\;=\;\ln\!\Bigl[\tfrac{27n^2 - 2}{27n^2}\Bigr].$
$\tfrac{27n^2 - 2}{27n^2}
\;=\;1 \;-\;\tfrac{2}{27n^2}$

$\ln\bigl(27n^3 - 2n\bigr)\;-\;3\,\ln(3n)
\;=\;
\ln\!\Bigl(1 - \tfrac{2}{27\,n^2}\Bigr)$
$\sum_{n=1}^\infty \ln\!\Bigl(1 - \tfrac{2}{27\,n^2}\Bigr)$
19:48
That looks to be basically correct but, if I were your instructor, I would not be very happy with the exposition. It is not very linear, and needs some cleanup.
But you can then express the sum of the logarithms as the log of a product, n'est-ce pas?
Ah okay
$\sum_{n=1}^\infty \ln\!\Bigl(1 - \tfrac{2}{27n^2}\Bigr)
\;=\;
\ln\!\Bigl[\prod_{n=1}^\infty \Bigl(1 - \tfrac{2}{27n^2}\Bigr)\Bigr]$
$\sin(\pi x)
\;=\;
\pi\,x \,\prod_{n=1}^\infty
\Bigl(1 - \tfrac{x^2}{n^2}\Bigr)$
where did you find this exercise?
Well, I don't know about that product off the top of my head, but assuming that this representation of $\pi$ is correct, it seems that you are done.
Since you can rewrite $\frac{2}{27n^2} = \sqrt{2/27}^2/n^2$.
$\sum_{n=1}^\infty
\Bigl[\ln(27n^3 - 2n) - 3 \ln(3n)\Bigr]
\;=\;
\ln\!\Bigl[\prod_{n=1}^\infty \Bigl(1 - \tfrac{2}{27n^2}\Bigr)\Bigr]
\;=\;
\ln\!\Bigl[
\frac{\sin\bigl(\pi \sqrt{2/27}\bigr)}{\pi \,\sqrt{2/27}}
\Bigr]$
@SineoftheTime they are exercises of analysis 1
@XanderHenderson Now what should I do?
I don't think so
from which book?
19:58
Pdf
@Binky What is the exercise? Again, assuming that the product representation of $\pi$ is correct, it seems to me that you are done, no?
@Binky yeah? which book?
pdf of what?
They are exam texts
00:00 - 20:0020:00 - 23:00

« first day (5259 days earlier)      last day (57 days later) »