« first day (2860 days earlier)      last day (2458 days later) » 

15:02
for normal problems it doesnt take years
@Liad Suppose $\pi_1(B)$ and $\pi_2(B)$ are both finite and find a contradiction
What do you know about $\pi_1(B)\times \pi_2(B)$ for example?
@AlessandroCodenotti i cant use the fact that the cardinality of sets is linear
@Arrow not sure that answers your question, but one way to think of the group (or more generally monoid) algebra is this: suppose $R$ is commutative (you want that to talk about $R$-algebras anyway, I think), then the free functor $\mathbf{Set} \to R\textrm{-}\mathbf{Mod}$ is strong monoidal where we consider the monoidal structure on $\mathbf{Set}$ given by products and on $R\textrm{-}\mathbf{Mod}$ by tensor products,
since there is a natural bijection $R^{(X)} \otimes_R R^{(Y)} \cong R^{(X \times Y)}$ satisfying the necessary coherence axioms.
@Abr001am If you can do such a problem under a minute, I think you master the work with telescoping sums to some good extent. math.stackexchange.com/questions/2805416/…
What, which order?
Sure, but that's not needed here
15:03
@Abr001am I mean to be able to do it naturally, without someone to tell you how to do it.
People don't upvote for the simple reason they don't see the simple transformation I did there.
(I told you, much experience is needed)
@AlessandroCodenotti ok i got what you want me to do. thanks.
You can conclude quickly from there
@MatheinBoulomenos I'm still studying this and I just feel like there are so many unmotivated names
yea because $B$ contained in the set you worte
@LeakyNun what do you mean?
15:06
like "complete extension" in def 2.4
like they just introduce a new type of extension each page
and I've no idea what the point is
@Liad indeed and now you're done
@Waiting can i ask you a question ? (about integrals i mean)
do we know that it will be countably infinite? @AlessandroCodenotti
@Abr001am Go ahead.
@LeakyNun yeah, it doesn't have much in terms of motivation
which isn't that surprising if you prove LCFT in under 30 pages
15:07
hmm...
@Liad yes, projecting can't increase the cardinality so $\pi_i(B)$ are at most countable and the sams holds for their union
right. nice , thanks
@MatheinBoulomenos :c
@waiting Ok if we have $f(x)=\int ln(x) dx$ if we integrate it by parts we get $f(x)=xln(x)-1 $ right ?
@Abr001am +C
15:10
+c , let it be. so it's correct.
I'm struggling with a proposition in chapter one of Neukirch @Mathei, I'm sure I'm missing something stupid, do you have time to help me with it?
No, it's not correct.
@Abr001am xlog(x)-x+C
@MatheinBoulomenos could you motivate things?
+d
@Waiting oh yeah i forgot to integrate the constant. that's true
15:11
@AlessandroCodenotti is doing LCFT?
@Arrow once you have the adjunction between monoids and $R$-algebras, you can use that to construct the arrow $RG \to \operatorname{Hom}_{R\textrm{-}\mathbf{Mod}}(R^{(X)},R^{(X)})$ from an arrow of monoids $G \to \operatorname{Hom}_{\mathbf{Set}}(X,X)$:
Just take the compose $G \to \operatorname{Hom}_{\mathbf{Set}}(X,X)$ with the monoid homomorphism induced on endomorphism monoids by the free module functor $\operatorname{Hom}_{\mathbf{Set}}(X,X) \to \operatorname{Hom}_{R\textrm{-}\mathbf{Set}}(R^{(X)},R^{(X)})$ and then apply the adjunction to get an $R$-algebra morphism $RG \to \operatorn
I'm writing with difficulty from the phone, @LeakyNun
@AlessandroCodenotti okay, I can try
@LeakyNun what exactly?
@MatheinBoulomenos complete extensions, finitely ramified extensions, complete unramified extensions, ...
like ok there are a lot of types of extensions
aber warum
@Abr001am you wrote $x\log(x)-1$ which is not correct (not $1$, but $x$).
15:13
@MatheinBoulomenos warum soll man care about dieser extensionem
please dieses Denglish ist killing me
Suppose $L/K$ is normal field extention and denote $H = Aut (L/K)$.
i need to prove that $L/L \ ^ H$ is separable and deduce that it is Galois.
Someone can help?
@LeakyNun LCFT?
@AlessandroCodenotti local class field theory
15:14
I have no idea what that is
isn't Neukirch about LCFT... whatever
@LeakyNun it's about ANT in general
@Liad definition of separable?
now, if let $u=ln(x)$ and calculate $f(x)$ we get $\int u e^u dx $ right @waiting which is $xe^x-e^x+C$ by parts.
@LeakyNun every element is separable
15:15
including CFT (local and global) in the later chapters
There is a chapter on it, but there are other 6 chapters about ANT as well
@LeakyNun infinite algebraic extensions of local fields are never complete, that's why you want to pass to the completion sometimes if you work with infinite extensions. This will still contain the original extension as a dense subfield
Finitely ramified extensions are nice, because then the ring of integers is still a DVR, that goes wrong with infinite ramification
@AlessandroCodenotti so what's the proposition?
@Abr001am Then?
@Mathei I'm looking at proposition 10.1, the one describing the factorization of $p$ prime in $\Bbb Q(\zeta)$
@Waiting if it's true that means an integral can be calculated in more than two ways yielding to completely different results .
15:19
$\zeta$ is a $n$-th root of unity so by proposition 8.3 we're interested in how $\phi_n(X)$ factors mod $p$
@MatheinBoulomenos is it easy to prove that the ring of integers is still DVR?
@Abr001am you made a variable change, remember? Return back in the final answer using the variable change you used.
@LeakyNun yes
Thus, you perfectly get that $int \log(x) \textrm{d}x= x\log(x)-x+C.$
@MatheinBoulomenos is there a 1-1 correspondence { finitely ramified extensions } <-> { complete extensions } that sends E to E-hat forward and E-hat to (E-hat intersect K^sep) backward?
15:22
@LeakyNun yes
When he says "observing that $f_p$ is the smallest positive integer such that $p^{f_p}\equiv 1$ mod $m$ it is obvious that..." I don't see why is that obvious
@MatheinBoulomenos could you give me some directions?
@AlessandroCodenotti f_p?
@LeakyNun it's all in lemma 2.2
Time to leave. Back a bit later. @Abr001am do some practice with simple examples and see how things work.
Piece of cake.
@MatheinBoulomenos that skipped a lot of steps
15:25
@MatheinBoulomenos thanks for the response and sorry for the late reply. That is a nice viewpoint but I just wanted to write the explicit procedure by which adjunctions give the correspondence. What I missed was the observation that the arrow assignments of the free-forgetful module adjunctions, when looking at a single object, are monoid arrow (by functoriality). From there it is clear :)
Another neat viewpoint related to yours is just to view reps and actions as functors from $\mathbf B G$ and the post-compose with the free/forgetful module functors to move between worlds.
@LeakyNun the important step is Prop 7.1 in the appendix which has a reasonably detailed proof
@AlessandroCodenotti so we're looking at the splitting field of (the reduction of $\Phi_m$ over $\Bbb F_p$). We know actually that $\Bbb F_{p^n}$ has a cyclic multiplicative group of $p^n-1$ elements, so it is actually the splitting field of $\Phi_{p^n-1}$ (in other words, every extension of finite fields is a cyclotomic extension)
@Waiting take care , i was just making certitude of an answer of mine.
@MatheinBoulomenos what the hell is a formal group?
(I know what its definition is, but what is the story behind?)
@LeakyNun the definition is in the paper
quite intuitive, right?
I've no idea why you treat a formal power series as a group
wiki says it stems from Lie group
15:33
@LeakyNun this has some motivation
it's related to elliptic curves
motivation!
@Arrow glad to help
@MatheinBoulomenos I agree with that but I still don't see how we reduce to the case $p\nmid n$
Ahhh, wait, I got it
I was mixing up the primes in $n=\prod p^{v_p}$ and this prime $p$ which is arbitrary instead
@MatheinBoulomenos great, another 25 pages to read
yeah, mod $p$ you just have $\Phi_{pn}=\Phi_n^p$
15:39
Isn't that $\Phi_n^{p-1}$? Or more generally $\Phi_{p^sm}=\Phi_m^{\varphi(p^s)}$ mod $p$?
@MatheinBoulomenos the notation for the multiplicative formal group law looks like the rank 1 split torus..
@AlessandroCodenotti oops, you're right. Well, all we need here ist that it is some power
@MatheinBoulomenos what is the intuition behind having the group law look like X+Y?
@MatheinBoulomenos Right, thanks! In the step before this one, where he goes from the congruence mod $\mathfrak p$ to the congruence mod $p$ that's just the inclusion $\Bbb Z/p\Bbb Z[X]\hookrightarrow\mathcal{O}_K/\mathfrak p[X]$ at work, right?
proofwiki.org/wiki/User:GFauxPas/Sandbox hope I didn't bite off more than I can chew with this one
took me a long time to even write the exposition
15:43
@AlessandroCodenotti yes
@GFauxPas thank you
Ok, thanks a lot!
what did I do
That's kinda like building the Borel hierarchy @GFauxPas
@LeakyNun so basically a group law is related to some p-adic analog of a one-paramter subgroup of a real Lie group $G$, i.e. a Lie group homomorphism $\Bbb R \to G$ where $\Bbb R$ has the Lie group structure given by addition. That's where the addition comes from
15:47
I know what a one-parameter subgroup is, but I'm not seeing anything
@LeakyNun so every Lie group has a compatible analytic structure, then for a one-parameter subgroup $H$, the group operation $H \times H \to H$ will be an analytic map, so locally it looks like a power series and since this is a one-parameter subgroup, it locally looks like addition
@MatheinBoulomenos why must a complete unramified extension contain an arithmetic Frobenius?
@MatheinBoulomenos I see
@AlessandroCodenotti intuitively, it's "keep on taking countable infinite unions of countable sets infinitely many times. Okay but you're not done yet"
but yeah Borel sets are sigma algebras
@LeakyNun for Galois extensions you always have a surjective map to the Galois group of the residue field. If the extension is unramified, this will actually be injective, too, so there's a unique lift of the Frobenius in the residue field to the extension. If you take the completion, you can extend by a density argument
you mean unramified
15:54
yeah
is your first sentence easy to prove?
ok I see that 2.5 basically defined your first sentence to be true
and then proved equivalence for finite extensions
@GFauxPas you pointed out that i said something mean to Kasmir without me noticing, i forgot to thank you.
@GFauxPas you also need complements or intersections, but yes that's the idea
right
oh no problem
@MatheinBoulomenos is this part important?
15:58
which part? The Frobenius certainly is important
@LeakyNun I think I already gave you a reference in Neukirch for the surjectivity of that map
yesterday, by MatheinBoulomenos
@LeakyNun the proof of that surjectivity is Thm. 9.4 in chapter 1 of Neukirch
it's important that lifting a Frobenius for unramified primes also works in the global case
but unless you choose a prime in the extension above the base prime, the Frobenius is only well-defined up to conjugation
(in the local setting, you always just have one prime, so this is no problem)
and of course for Abelian extensions this is no problem, which is used in global CFT
I see
16:15
$ \substack{ \\ \\ \Huge{\text{K}} \\ {n \ge 0 }} a_n $
Ma LaTeX skills r gettin' off da charts, y'all.
oh my
@LeakyNun oh I somehow missed the question why the ring of integers in a finitely ramified extension is still a DVR
the thing is you have a unique extension of the valuation and if you have finite ramification, this is still a discrete valuation
The image of the extension of a valuation $v$ on $K$ to some extension $L/K$ which I will also denote by $v$ is given by $v(L^\times)=\frac{1}{e(L/K)} v(K^\times)$
unramified extensions don't change the image of the valuation (this is also called the value group of the valuation), even infinite ones
0
A: Fixed Fields and normal extensions

Stefan4024For the first part take any $\alpha \in L$. Then let $\{\alpha_1,\dots \alpha_n\}$ be the set of distinct elements obtained by $\text{Aut}(L/K)$ acting on $\alpha$. Note that this set is finite, as the extension is algebraic. Now consider: $$h(x) = \Pi_{i=1}^{n} (x-\alpha_i)$$ Now it's not hard...

stefan wrote in the answer " Note that this set is finite, as the extension is algebraic. .."
is it true?
16:31
@MatheinBoulomenos why is that? there are not finite algebraic extentions
any element obtained from $\alpha$ by an automorphism that fixes $K$ is still a root of the minimal polynomial of $\alpha$ and that has only finitely many roots
$Aut(L/K)$ could be infinite right?
that's not an issue
but yeah
and why does $h$ is fixed by $Aut(L/K)$ ? i see why all the $\alpha_i $ goes to $\alpha_j$ but we are left with $\sigma(x)$
$\sigma \in Aut(L/K)$
$\sigma$ just acts on the coefficients of the polynomial
16:34
didnt he mean that $\sigma(h(x) ) = h(x)$ ?
$h(x)$ is a polynomial
how do you define $\sigma(h(x))$ if not by letting $\sigma$ act coefficient-wise?
kk got it ^^
thanks
proofwiki.org/wiki/User:GFauxPas/Sandbox what do you guys think about the Question I have on this theorem
@BalarkaSen Hey, congrats on clearing ISI written test :) .. and did you receive CMI admission offer?
in terms of exposition, just to help the reader understand
not a question about the math involved, its a question on how to present the theorem in a clear way
16:48
@LeakyNun, Let $f:[-1,1]\to\Bbb R$ defined by $\sin x$ if $x\in[-1,1]\cap\Bbb Q$ and $0$ otherwise. Then, $f$ is not Riemann integrable in any interval containing $0$, right?
right
thanks!
@MatheinBoulomenos $\{\alpha_1,\dots, \alpha_n\}$ are the roots of the minimal polynomial of $\alpha$ ?
@Liad yeah
over $K$ ?
or $L \ ^ H$ ? does it matter?
16:53
is it ok to ask a math question here?
over $L^H$. The coefficients don't have to lie in $K$
oh ok. Me and friend talked today about Brownian motion and ergodicity - does every Brownian motion is ergodic?
is an ergodic process*
@Myprofileissaved no, you're not allowed to talk about math here
Lol mathein
@Myprofileissaved long as it doesnt have numbers you should be ok
these guys rarely do any math with numbers in it
17:06
what's a math
@MatheinBoulomenos In definition 3.3 of this, why $\mathcal O \subseteq \Theta^L_{\pi,\pi}$?
@Faust let $K$ be number field and let $x \in K$. There's your number: $x$
thats exactly what i am saying the number is always a variable
O.o
Fix $x \in K$
lol
17:09
@MatheinBoulomenos why does $L \ ^ H / K$ is normal?
If I catch any of you nerds doing math...
@LeakyNun the definition simplies to just elements in $\mathcal{O}_L$ which are fixed by the Frobenius when both uniformizers are the same
why are elements in OL fixed by Frobenius?
no, $\mathcal{O}$ is the ring of integers in $K$
the elements in the base field are fixed
by Galois theory
the ring of integers of an extension is also fixed by the galois group isn't it?
17:11
@ÍgjøgnumMeg not pointwise
or am I talking out of my a*se
oh I see
I also
just failed at censoring
おめでとう
Let $A$ be the the closure in $\Bbb R$ of the set $\{n+\frac1m+\frac1k:n,m,k\in\Bbb N\}.$ So, $\Bbb N\cup \{n+\frac1m:n,m\in\Bbb N\}\cup \{n+\frac1m+\frac1k:n,m,k\in\Bbb N\}\subset A'$ but $\Bbb N\cup \{n+\frac1m+\frac1k+\frac1l:n,m,k,l\in\Bbb N\}\nsubseteq A'$ where $A'$ is limit point of $A$. Am i right?
@MatheinBoulomenos in Lemma 3.4, how can f satisfy those conditions?
[PhD]ok... I cannot manually click through 5376 files!
17:17
hi @BalarkaSen
@LeakyNun you can take for example $\pi X + X^q$
how is that X^q mod p?
oh
@Sawarnik Thanks, but I really can't be any less invested about all this admission garbage. (To answer the second question: No, I did not. I firmly believe I won't make it to CMI.)
ich bin ein Esel
@MatheinBoulomenos so basically f = pi X + X^q + pi X^2 g?
17:20
Hi @LeakyNun
@MatheinBoulomenos what a terrible way of saying it
@Faust I use numbers all the time. I use $\exp 0$, $0$,$1$, and $2$
And $\pi$
Are there other numbers?
a couple more...
But, like, useful ones?
hmm
17:25
Hmm really makes you think
yeah i got nothing
@MatheinBoulomenos, will you please check this:
14 mins ago, by Silent
Let $A$ be the the closure in $\Bbb R$ of the set $\{n+\frac1m+\frac1k:n,m,k\in\Bbb N\}.$ So, $\Bbb N\cup \{n+\frac1m:n,m\in\Bbb N\}\cup \{n+\frac1m+\frac1k:n,m,k\in\Bbb N\}\subset A'$ but $\Bbb N\cup \{n+\frac1m+\frac1k+\frac1l:n,m,k,l\in\Bbb N\}\nsubseteq A'$ where $A'$ is limit point of $A$. Am i right?
17:40
@LeakyNun, please!
17:59
$A'$ is $A$ minus isolated points, or something like that, right?
@AkivaWeinberger yes
Limit points of A
I think $A'$ is $\overline{\{n+\frac1m:n,m\in\Bbb N\}}$
And then $A''$ is $\Bbb N$
unless I made a mistake somewhere
@AkivaWeinberger ok! thank you very much
::Reading order theory while my files are uploading::
18:25
Can anyone help out ? ... What would be the argument of $i^{-7}$ ?
I'm getting $\frac{\pi}{2}$ ...
But my book says $\frac{3\pi}{2}$ ...
@BalarkaSen I know you don't care but I wanted to know you'll go anyway, especially since we are going to the same institute.
@Sawarnik Oh yeah sorry for not congratulating on my part. The reason for "not investing" is not pretentious, it's just that the admissions have been taking a toll on my sanity lately and I don't want to think about it a lot
How's life?
Henlo
Hey
(Check out my last message on DC)
18:50
If $\dim_{A/\mathfrak m}(A/\mathfrak m^2)$ is known, where $A$ is a local ring and $\mathfrak m$ its maximal idea, then is $\dim_{A/\mathfrak m}(A/\mathfrak m^n)$ known?
No that makes no sense
@BalarkaSen lfmao
19:16
What is the usual definition of the image of a point under a function which is not inside the domain of the function? Is it the empty set or rather undefined?
Let $H$ be a Hilbert space, $T \in B(H)$, and $x,y \in H$ unit vectors. I am trying to show that $\langle Tx,y \rangle^2 = |\langle Tx,Ty\rangle |$. The best I can do is: $$\langle Tx,y \rangle^2 = \langle Tx,y \rangle \cdot \langle Tx,y \rangle = \langle T\langle Tx,y \rangle x,y \rangle$$...which isn't very much...I could use a hint.
I think I need to show that $T\langle Tx,y \rangle x = T^*Tx$, but I don't know how to do this.
suppose $\{\sigma(\alpha) : \sigma \in Gal(K/F) \}$ is a normal basis for $K/F$. i need to prove that $K = F(\alpha)$.
so i wrote the basis as $\{\alpha_i\}$
where the $\alpha_i$s are the roots of the minimal polynomial of $\alpha$. i need to show that $\alpha_i \in F(\alpha)$,
someone see how?
19:34
Consider an involution in $S_n$. What is the cardinality of its centralizer?
I've forgotten most of the Galois theory I ever knew. What's a normal basis? @Liad
@Nûr Of or in?
In
:)
@philmcole In what situation did this come up?
@Nûr And centralized is the set of things that commutes with it?
@AkivaWeinberger Discussing maps on submanifolds
I would imagine the empty set, but I think if $A$ isn't completely contained in the domain of $f$ I'd write $f(A\cap\operatorname{dom}f)$ rather than $f(A)$
@Nûr I know that it's even
19:38
Yes, the commutant
@AkivaWeinberger Ok thanks. I would have written the concrete part of the text where this came up in, but it is rather long and also in German. I also think it makes more sense to be the empty set in this case.
Do you have a way to visualize how a complex linear function acts? With real ones it is easy with spectral theorem + polar decomposition.
$\Bbb C^n\to\Bbb C^n$ or $\Bbb C\to\Bbb C$? 'Cause the latter is just a dilation and rotation
$\Bbb C^n\to\Bbb C^p$ I meant the field was $\mathbb C$
You stop thinking about how it acts visually and start thinking about the algebra
I think
19:50
lol
I say this as a topologist
So you know it pains me
@MikeMiller I can't help but imagine Mathei in the corner laughing evilly
When the field is $\Bbb R$. spectral theorem + polar decomposition imply the linear function acts on a certain basis as a rotation then as a scaling for each vectons of the basis.
@LeakyNun are you here please?
You could also do JNF
Which says you merely need to understand dilation, rotation, and "complex shearings"
Which I have no picture of personally
19:55
if $d(f(x),f(y))<d(x,y)$ and $x_{n+1}=f^{n+1}(x_0)$ if for example i want to to see $\lim_{n\to+\infty} d(x_n,x_{n+1})$ @LeakyNun
@LeakyNun $d(x_n,x_{n+1})= d(f(x_{n-1}),f(x_n))<d(x_{n-1},x_n)<...<d(x_0,x_1)$ then the limit is not 0
how to prove that it is a Cauchy sequence ?
:D
I think last year my understanding of the tangent and cotangent bundle was held back because I wanted to see them visually, I feel like I understand them much better now after we saw them in class and just use them as the natural spaces to define differentials and stuff without ever drawing a picture
The tangent bundle deserves a picture (directions to move in)
The cotangent bundle deserves algebra (ways to eat the directions you move in)

« first day (2860 days earlier)      last day (2458 days later) »