« first day (3768 days earlier)      last day (1246 days later) » 
00:00 - 18:0018:00 - 00:00

12:43 AM
derivative of projection operator, $\pi:R^{m+n}\to \mathbb {R^n}$ at $x\inR^{m+n}$ is $\pi$. right?
since $\pi$ is a linear transformation.
 
yes
 
Sanity check: so if I have a function G(t,v) such that G(t,V_T) = 0 for all t \geq 0. Then the partial derivative of G with respect to t at position (t,V_T) must be zero right
Because in an exercise, I need to take the laplace transform of the partial derivative of G, with above conditions, and show that the result is one
i.e. I need to show that $\int_0^\infty -\partial_t G_t(t,V_T)e^{-st}\, \mathrm{d}t = 1$
 
@Thorgott Thank you very much
 
@user2103480 $V_T$ is?
 
some random number
not random in probability, literally just some constant
 
12:56 AM
then yes
youre sane
 
so the statement must be wrong duh
 
why are you writing $G_t$
 
G_t for time derivative of G isn't that uncommon of notation
 
because I'm stupid
 
oh. one or the other, yeah
grading HW blues: so many solutions with the same error leading to the wrong answer....which is the same typo as in the solutions manual :|
 
1:00 AM
hahahah classic
 
1:10 AM
yeah ok, then I don't see why that shouldn't be zero
 
Just counted: 40 / 81 students did it
yaaaaay /s
 
oof
 
@Thorgott me neither
gonna write the guy
 
@Thorgott amusingly, this statement seems appropriate to either conversation
 
true
 
2:07 AM
Can someone take a look at my question here?
0
Q: Does the direct sum decomposition of a p-group tell you the number of Prüfer subgroups?

Keshav SrinivasanMy understanding is that every countable Abelian $p$-group can be written as a direct sum of at most countably many finite cyclic groups and at most countably many copies of the Prüfer $p$-group. Suppose that when such a direct-sum decomposition is written for a particular countable Abelian $p$-g...

I need the answer to it in order to make progress on a research problem I’m working on.
 
2:20 AM
the context obscures what's going on here
can you find more than $2$ isomorphic copies of $G$ in $G\times G$ where $G$ is a non-trivial group?
 
2:35 AM
How does a homogenous degree-1 polynomial in n+1 variables induce a section of the tautological line bundle over $\mathbb{CP}^n$?
 
@feynhat please clarify. Degree 1? These are the sections of the hyperplane bundle, dual to tautological.
 
Yes. I meant dual of tautological bundle.
 
So show that any linear homogeneous polynomial gives a linear functional on the lines.
 
2:56 AM
Suppose the polynomial is F, and $\pi^{-1}(l)$ is the fiber over $l \in \mathbb{CP}^n$. Define a functional $\pi^{-1}(l) \to \mathbb{C}$ as $v \mapsto F(v)$. This works?
 
Yes. You actually want a global definition, not a fiberwise definition, no?
 
So I need to check this is holomorphic as $l$ varies, right?
 
Don’t forget how $F$ is defined in the first place.
 
3:15 AM
x''(t)=f(x(t)) is given. what is equillibrium? potential?classifying equillibrium. where do I get the theory of this question? Can you suggest some book/ link?
 
I mean F itself is holomorphic, so the map $l \mapsto (v \mapsto F(v))$ is holomorphic.
 
@Semiclassical so does that mean 40 out of 81 students rote memorized the solutions manual with the typo?
 
Rote memorization in math? That's a funny thought
 
Perhaps, include that question on a quiz to see if the same typo shows up?
@Rithaniel physics, actually.
 
Ah, that meakes slightly more sense, but is still mostly funny
 
3:29 AM
Indeed.
 
$x''(t)=f(x(t))$ is given. what is equillibrium? potential?classifying equillibrium. where do I get the theory of this question? Can you suggest some book/ link?
here I made equation Tex form.
 
3:52 AM
@skullpatrol no? they can download it online
it's not hard to track down, unfortunately. it's a well-known textbook
that said, the typo is actually just a math one
the physics that's written down is fine, but the simplified form they then write down is wrong
that said, I dislike the problem from the start because the solution comes down to "solve a cubic equation numerically"
which is gross, because that's not something you can ask for on a quiz
 
4:51 AM
That's what I meant, they copied it down without thinking about the the math involved in the simplified form @Semiclassical; but, since you can't put it on a test, it doesn't really matter :-)
 
user486313
sup
 
user486313
@skullpatrol you're everywhere :)
 
chillin' how are you?
 
user486313
@skullpatrol should I do Algebra or Analysis?
 
user486313
@skullpatrol life is short, please tell me
 
4:55 AM
Start with your strong subject
 
user486313
I see ...
 
user486313
and then?
 
user486313
that's it, @skullpatrol?
 
user486313
:(
 
user486313
I want to be famous in Math
 
user486313
4:57 AM
like, literally, famous
 
and then work on your weak subjects
 
user486313
I see ...
 
user486313
I spoke with a renowned Mathematician tonight
 
and?
 
user486313
he's from Princeton, I think
 
user486313
4:58 AM
so, I gave him my idea
 
user486313
he said, rigorously, it's impossible to do, while numerically it's likely already been done, namely by the many engineers that probably have thought of what I've proposed to him.
 
user486313
@skullpatrol do you think I'm wasting my time doing Algebra?
 
What was his advice?
@user131585 no
 
user486313
@skullpatrol his advice was to find an expert that knows my question better, in order to understand the knowns and unknowns, of the state of knowledge. He only gave me his best guess.
 
user486313
oh, he did suggest two people to reach out to, so that was nice of him
 
5:02 AM
cool
 
user486313
@skullpatrol people say doing Algebra is a waste of time
 
user486313
I am scared
 
user486313
Hi @TedShifrin
 
@user131585 Can I ask what was your question to that Mathematician?
 
@user131585 Everything builds on Algebra.
 
5:03 AM
Hello All. I have a problem I've been struggling with. I think I have some intuition for the solution, however I can't be sure. My problem is as follows: given four observers in a 3-d space, and some signal source with unknown coordinates, and the velocity of said signal, I want to calculate both the minimum and maximum possible time of arrival differences (i.e. difference between time of arrival at any two observers).
I think the max would be at one (any?) of the four observers, but I could be wrong. Not sure about the min.
 
user486313
@skullpatrol But people don't / can't find jobs after their PhDs in Algebra
 
They specialize in other branches
 
user486313
@Spade000 it's a problem in fluid dynamics, on the stability of shear flows, given some initial data, for instance
 
user486313
@skullpatrol is there a particularly good branch to move into?
 
user486313
@skullpatrol my thesis advisor from the past is a number theorist and algebraic geometer
 
user486313
5:06 AM
she moved to the other side of the world, to keep an academic job ...
 
Do what you love doing.
 
user486313
True true ...
 
cya
 
user486313
bye :(
 
 
2 hours later…
7:46 AM
What is the splitting field of $x^4+2x^2+2$ over $\Bbb F_3$?
 
8:26 AM
oops
no it's not
 
Oh I think I can handle that
0
Q: composite of radical extensions is radical extension

love_sodam Let $E_1/F$ and $E_2/F$ be two radical extensions. Then $E_1E_2/F$ is also radical extension. Since $E_1$, $E_2$ are radical extensions over $F$, there is $F\subset F(\alpha_1)\subset\cdots F(\alpha_1,...,\alpha_n) =E_1$ such that $\alpha_1^{n_1}\in F$ and $\alpha_i^{n_i}\in F(\alpha_1,...,\alp...

Could anyone check my proof?
 
just show the polynomial is irreducible and then quotient out by it
 
 
2 hours later…
10:31 AM
If f=x^p-a is an irreducible polynomial in F[x] where char(F)\neq p and p is odd prime, then $\alpha\notin F(\alpha)-F(\alpha)^p$ where $\alpha$ is a root of $f$?
If $\alpha$ is contained, then I can't see where the contradiction arise
 
Let's saaaay I have a discrete valuation ring $R$ with residue field $k_R$ and I have a ring extension $R \to S$ such that $k_R = k_S$.. can I somehow get an isomorphism $k_S \otimes_{k_R} k_R^n \to S \otimes_{R} k_R^n$
because I want one
i mean the guy on the left is just an $n$-dimensional $k_S$-vector space so really I just want $S \otimes_R k_R^n$ to be an $n$-dimensional $k_S$-vector space but
idk
Here's the real sitch: let $K/\Bbb Q_p$ be an unramified extension and let $K^{ur}$ be the maximal unramified extension of $K$ and $\widehat{K^{ur}}$ its completion inside $\Bbb C_K$. Let $E := k/\mathfrak{p}_K$ and $E^s$ a separable closure of $E$. Then $G := \operatorname{Gal}(K^{ur}/K) \cong \operatorname{Gal}(E^s/E)$. Let $T$ be a finitely generated $\mathcal{O}_{\widehat{K^{ur}}}$-module with a continuous semilinear action of $G$. Suppose $pT = 0$ ($p$ being a uniformiser).
Then $T = T/pT$ is a discrete space so there's a continuous hom $\mathcal{O}_{\widehat{K^{ur}}} \otimes_{\mathcal{O}_K} T^G \to T$ (the natural hom) and $T$ is an $n$-dimensional $E^s$-vector space, so that $T^G$ is an $n$-dimensional $E$-vector space. Thus we have a hom $\mathcal{O}_{\widehat{K^{ur}}} \otimes_{\mathcal{O}_K} E^n \to (E^s)^n \to E^s \otimes_E E^n$
And I want $E^s \otimes_E E^n \to \mathcal{O}_{\widehat{K^{ur}}} \otimes_{\mathcal{O}_K} E^n$ to be an iso as well because then everything works out and I can stop crying
alas
 
I'll pray for you 🙏
 
this might all be complete nonsense btw
 
10:46 AM
too many symbols for me to read this late at night
 
alas
I'm like that scientist dog meme that says "I have no idea what I'm doing"
$E$ should be $\mathcal{O}_K/\mathfrak{p}_K$ lol
 
me too 🙏🏽
 
idk, finding an isomorphism that goes from the residue field to the ring like that seems unpossible
try to construct something explicitly and depending on how badly you fail you can uh, decide how much more to cry maybe
 
I'll just wait for Lukas to turn up and tell me I'm stupid
 
user image
5
 
10:57 AM
 
I spent way too much time on this
 
lmfao I didn't even notice
thanks for pointing it out
I thought it was just the generic meme
 
all my efforts would have been in vain
You would have missed this jewel of comedy
 
but lo, he did make it known to us that it was not so
 
11:14 AM
Question about MSE: I failed a review audit, but I am not convinced I was wrong. Is there a way to tell the mods that a particular question might be a bad example for a review audit? Or is there some other course of action I should take?
 
11:24 AM
tutor seems to think I'm done and then explained to me why I'm done and I still don't know why lmaooo
 
Good job!
 
12:05 PM
@EdwardEvans me
In my course on stochastic processes in fluids we were talking about hamiltonians all the time and I did not at all understand how the time dependence works
Turns out there is a thing called "hamiltonian dynamics"
(I know this is well known, just had no clue about the topic)
 
lecturer didnt care to mention even though he knows I'm not a physicist
it's just second nature to him hahaha it's not intentional he doesn't mention those things
 
what a nerd
tbf I feel like being a physicist would be cool
like when you tell people you're a mathematician they're scared by it and revere you as being really intelligent but when they ask what you do and you're like "yeah I try and find out if smth is prime" they're like "nerd"
but if you're a physicist and you go like "yeah I study atoms" or whatever the fuck physicists do they're like woahhh that's so coool
chad physicists vs virgin number theorist
not bitter
 
Just say you do the complicated part of physics
 
or start shouting about how physics and number theory are intricately related á la Edward Frenkel
and get carted off by the men in white coats
screaming "the digits have an end"
or however tha meme goes
 
12:15 PM
@EdwardEvans NUMBERS HAVE AN END
mind + logic
 
yeah, shout about how the laws of physics are invariant under change of field to justify talking about p-adic physics
 
@EdwardEvans nah mathematics is cooler
 
"I'm such a chad I don't even look at the real world"
 
you have pretty direct access to many scientific areas
including physics
 
I was interviewed for my scholarship by a panel of non-mathematicians and was trying to explain to them what Iwasawa Theory is
and they were like
"Why is it relevant to real world applications?"
 
12:17 PM
lmao
get rekt
 
and I replied "It literally isn't, in fact it's only barely relevant to mathematicians"
 
lol
 
and I still got the scholarship sooo
 
still got slammed hard
thrown across the room
 
absolutely decimated
 
12:19 PM
yeah I study constructible molecules as part of chemical galois theory
adding atoms is just field exts ez
 
lmao
 
Isn't there a chance Iwasawa theory lets us learn about primes -> applications to encryption ?
 
@Astyx number theorists just invent that whole encryption BS to post on reddit
 
with number theory you can just say "crypto" to palm off the haters
 
ya
 
12:20 PM
chad hardy embracing detached mathematics vs virgin modern number theorist babbling about encryption
 
and wot
Iwasawa theory is just a structure theory of 2-dimensional complete Noetherian regular local rings
 
number theory, AG and algebraic topology are just the trickle down economics of mathematics
 
lmao
 
ANT, AG, AT: The Holy Trinity
 
the forefront of abstract mathematics that give the other areas useful tools every few years (or decades)
 
12:23 PM
Doesn't AT have direct applications in data science ? with topological algorithms or whatever they're called?
 
algebraic statistics
 
AG can still cash in on the crypto memes too, just say elliptic curves
 
abelian variety crypto
 
@Merosity or do stochastic calculus, and cash in on the crypto itself
 
12:46 PM
@Astyx You're talking about this.
 
TOPOS
 
Hi Bal.
 
hi hat
rattling hi hats
 
@feynhat yes that's it
 
Mike says it's useless.
 
12:50 PM
it sounds cool
 
I am confused a little with this thing. Given a degree d, homogeneous polynomial F in n+1 variable, it 'induces a section of the d-th tensor power of the dual of the tautological line bundle of CP^n'. How? It seems easy for d=1, for any l in CP^n, define the section as s(l)(v) = F(v), right? How do I generalize this to higher degrees?
Lets do it for CP^1. Given F(x, y) = x^2 + y^2. What section does it induce on $(\gamma^*)^{\otimes 2} \to \mathbb{CP}^1$? ($\gamma$ is the tautological line bundle of $\mathbb{CP}^1$).
 
1:08 PM
@feynhat (x o x) + (y o y)
 
So your [x,y] is in CP^1 and the section takes that to $\pi_x \otimes \pi_x + \pi_y \otimes \pi_y$, where, $\pi_x$ is the projection to x, etc.
 
2:00 PM
If A, B satisfies hypothesis of going down (A, B are integral domains, A is integrally closed, B is integral over A), is it possible that if i is the inclusion map i: A -> B then i*: spec(B) -> spec(A) is not open?
 
@EdwardEvans you just need $E^s \otimes_E E^n \to \mathcal{O}_\widehat{K^{ur}} \otimes_{\mathcal O_K} E^n$ to be an isomorphism, right?
That's not hard. We have that $p$ is a uniformiser in $\mathcal{O}_{\widehat{K^{ur}}}$, so that $\mathfrak{p}\mathcal{O}_{\widehat{K^{ur}}}$ is the maximal ideal of $\mathcal{O}_{\widehat{K^{ur}}}$ and we have $$E^s \cong \mathcal{O}_{\widehat{K^{ur}}}/\mathfrak{p}\mathcal{O}_{\widehat{K^{ur}}}\cong \mathcal{O}_{\widehat{K^{ur}}}\otimes_{\mathcal{O}_K} \mathcal{O}_k/\mathfrak{p}=\mathcal{O}_{\widehat{K^{ur}}} \otimes_{\mathcal O_K} E$$
 
2:15 PM
@feynhat Probably, but the point is just that you already know how x determines a section of the tautological bundle, so x^n gets you a section of its nth tensor power.
 
2:35 PM
https://math.stackexchange.com/questions/3924843/if-xp-a-in-fx-has-alpha-as-a-root-then-alpha-in-f-alpha-setminus-f
Anyone who can solve this?
 
3:14 PM
Hello everyone. I'll be sending an email for a postdoc and I don't know how to address her.

Her title is assistant adjunct professor, so I'm not sure if "Hello Professor X" would be right
What is the proper way?
She's a postdoc at a relevant institution. I'll be sending her an email because she changed to math after a BSc in a non-math degree, and I identified with that. And she researches something that also interests me. Maybe ask for some advice?
 
 
1 hour later…
4:31 PM
What's wrong with "Hello Dr. XY"?
 
lol
 
@LukasHeger Damn, thank you man
 
And $G$ should act componentwise on $T \cong (E^s)^n$ right?
I mean, otherwise $T^G \cong E^n$ doesn't work
 
not sure about that
 
4:45 PM
yeah that's what I'm worried about lol
 
@user2103480 sup bro Doctor XY
 
Could someone translate (2) for me ?
 
you still get an isomorphism $T \cong E^s \otimes_E T^G$ that is $G$-equivariant
that is just from Galois descent
and the isomorphism $E^s\cong \mathcal{O}_{\widehat{K^{ur}}} \otimes_{\mathcal O_K}E$ is also $G$-equivariant
 
and then I just put the second isomorphism into the first product
 
so if you combine those, you get a $G$-equivariant isomorphism $T \cong E^s\otimes_E T^G \cong \mathcal O_{\widehat{K^{ur}}} \otimes_{\mathcal O_K} E \otimes_E T^G \cong \mathcal O_{\widehat{K^{ur}}} \otimes_{\mathcal O_K} T^G$
 
4:50 PM
yeah
 
which is what you want, I guess?
 
yeah that's exactly what I was looking for
 
great
 
this is an inductive argument for p^n torsion $\mathcal{O}_{\widehat{K^{ur}}}$-reps
and this is the p torsion case
you can see the pset on Mampf if you're interested :P
 
I see
 
4:52 PM
is pretty interesting what we're doing
although Venjakob's handwriting is illegible
 
is there any field over which we know the inverse Galois problem to be true?
 
C(t) man
 
show it by pure algebra then
 
by pure topology
any fg group is quotient of a free group $F_n$ by a subgroup $H$; delete $n+1$ points from $\Bbb C$ and construct the cover corresponding to the subgroup $H$ of $\pi_1(\Bbb C \setminus \{n+1\, pts\})$
this cover has deck transformation group your favorite fg group
the cover is realizable as a Riemann surface by blackbox
the deck transformation group is the Galois group
 
why is any fg group quotient ?
 
5:05 PM
by definition of being fg; take a finite generating set
 
isomorphism theorem
 
look at the free group generated by them; look at the normal closure of the subgroup generated by the relators
thats $F_n$ and $H$
 
Ah right
Nice
 
who cares about normal closure, take the hom $F_n\rightarrow G$ sending the generators to the generators and then do isomorphism theorem
ill come back to this once my riemann surfaces lecture has progressed further
 
F(\alpha)^p is not a field if char(F) is not p right?
 
5:15 PM
what's that?
set of all $p$-th powers in a given field?
 
Yes
F^p = {a^p: a\in F}
 
consider $\mathbb{R}^3$
 
For general F
 
what does that mean?
the answer won't be the same for all F
 
I mean for general field F, it not ture right?
 
5:19 PM
he just gave a counterexample to your statement
 
yea I know
 
I don't know what a general field is
 
I was trying to solve that if f=x^p-a\in F[x] with char(F)\neq p where p is an odd prime, \alpha\notin F(\alpha)-F(\alpha)^p where \alpha is a root of f
seems much harder than I thought
 
what you write would be much more readable if you put $ signs around the math
 
I was trying to solve that if $f=x^p-a\in F[x]$ with $char(F)\neq p$ where $p$ is an odd prime, $\alpha\notin F(\alpha)-F(\alpha)^p$ where $\alpha$ is a root of $f$
sorry I thought without $ it would be more readable
 
5:25 PM
in case you are unaware, you can get latex to work in chat
 
alright
 
are you supposing $f$ is irreducible
 
Yes $f$ is irreducible forgot to type in
 
5:41 PM
oh ffs. the solutions manual for the textbook I'm grading HW on is available via fkn google books
 
Good evening. I have a math question, but I don't know how to properly tag and phrase it. Could someone please make suggestions?
 
what is the question?
 
as a general rule, it's best to give the question first and then ask for help
 
@Semiclassical Oh hi
Struggling with my EM assignment lol
 
what's the HW on
 
5:44 PM
How do I represent a "choice" in a formula?
so say I have a function f(x,y), where y is a value between 1 and 6. If y < 4, f(x) = x*0.8, and if y >= 4, f(x) = x*0.4.
 
@MechMK1 that's called a piecewise function
typically you'd write it like this:
 
there's two questions which just gives some physical configuration and asks me to compute the magnetic field/potential and one which seems to be a dipole construction
 
$$f(x,y)=\left\{\begin{array}{ll} 0.8x, & y<4 \\ 0.4x, & y\geq 4\end{array}\right.$$
 
dunno how to be more specific that that
 
so it gives you the source and asks you to compute the fields
 
5:46 PM
@Semiclassical I assume that's not supposed to render in chat?
 
not by default. but you can use the "LaTeX in chat" link (upper right)
 
yeah something like that
 
@MechMK1 in the description is a link to ChatJax. If you install that, you should see it render.
 
@BalarkaSen so like, a ring of current or something
 
@MechMK1 Yes. LaTeX in chat link
 
5:47 PM
(that's the usual Ampere version of a magnetic dipole)
 
@Thorgott I found that if $\alpha\in F(\alpha)^p$ then for some $\beta\in F(\alpha)$, $\beta^p=\alpha$. Then, $F(\alpha)=F(\beta)$. Is this shows something?
 
Ah, I see. I just googled what piecewise functions look like, that makes sense. Does it still make sense if I ask the question on the main site? Or would that be too broad or "too basic"?
 
oh so the first one is like, you have a setup like this: --O-- where the current comes in, goes across the sphere, goes out
 
it's not too broad, but it's probably already been asked/answered before. i'd bet there's a whole "piecewise functions" tag on the main site
 
True. Anyways, thank you very much for your help in this.
I hope you all have a nice weekend
 
5:50 PM
you too, happy thanksgiving
@BalarkaSen gotcha. so a surface current problem
 
yeah
 
Ah yes, it's thanksgiving in the US. Then happy thanksgiving as well to all those who celebrate it
 
that's tedious yeah
are you supposed to compute the field everywhere or just in a specific domain
 
everywhere haha
 
gross
 
5:51 PM
i know
 
I'm not even sure that has a closed-form answer. something something elliptic integrals maybe
 
all the instructors problems have been like this. the last one made us solve a Poisson problem with 18 boundary conditions
yeah people are getting weird integrals lol
 
for the simpler case of a single loop
so this one is presumably even worse
 
Lmao
 
that said, you probably want to start by doing the vector potential
because then you can take curl to get the magnetic field, rather than doing another horrible integral
 
5:57 PM
yeah
that sounds right
 
also lets you avoid having to use the Biot-Savart law which is horrible
in favor of a triple integral. still gross but not quite so bad
 
00:00 - 18:0018:00 - 00:00

« first day (3768 days earlier)      last day (1246 days later) »