« first day (1357 days earlier)      last day (3664 days later) » 

10:03 PM
$\phi\left(3^{m+2}\right)=2\cdot3^{m+1}$, but I guess the exponents are too small.
 
@seaturtles Seems so.
 
10:20 PM
@r9m did you see these ones?
@r9m I employed the same trick as above. This is the natural approach for these questions.
 
r9m
@Chris'ssis yas :D .. nice :)
 
@r9m Any limit like that involving $1/2$ should make us think of symmetry. :-)
 
r9m
@Chris'ssis yes ..
 
A thought bothers me though ....
@r9m I wonder how I could successfully use the symmetry to fastly finish such an integral. This result I proved some weeks ago and it's my best achievement so far (this year).
 
r9m
@Chris'ssis WoW .. !! that looks awesome .. !!
 
10:33 PM
@Chris'ssis doesn't the general answer handle those?
 
@robjohn there I used the symmetry $x_i\mapsto 1-x_i$
@r9m I agree. I worked some days (2 or 3?) on it and suffered a lot ...
 
@Chris'ssis I posted an answer using the CLT and that should work for both of those.
@Chris'ssis at least it gives the same answer :-)
 
@robjohn Yeah, CLT should work without any problem.
 
Is there not an answer for f(n) ? already?
 
@N3buchadnezzar several
@N3buchadnezzar for example: math.stackexchange.com/a/764938
 
10:35 PM
$$ \lim_{n\to\infty}\int_{[0,1]^n} f\left( \frac{x_1 + \cdots + x_n}{n} \right) \, \mathrm{d}x_1 \cdots \mathrm{d}x_n= f\left(\frac12\right) $$
Or something
 
@r9m It's the most beautiful thing I attended this year, something divine that originates from heaven. :-)
@r9m I attended a lot of series and integrals proposed by Ramanujan, but nothing compares to this one.
There are no words to talk about it, it's beyond words.
 
r9m
@Chris'ssis the value of the integral involves $\zeta$ .. $\zeta$ scares r9m
 
@r9m There are far worse functions to work with
 
r9m
@N3buchadnezzar then imagine what worse functions do to me .. :'(
 
Like tax returns and child support function papers
 
10:47 PM
like the function $3$
 
@Mike hai
 
@Mike Yes.
 
@Pedro look at the one with 62 upvoted.
 
@Mike Yeah, that was anon.
=D
 
10:56 PM
Different anon.
 
ORLY.
True dat.
 
If that was our anon I would have to beat him to death.
 
why is $\mathbb{Z}[\sqrt{n}]$ the integral closure of $\mathbb{Z}$ in $\mathbb{Q}[\sqrt{n}]$?
it could be relevant that $n$ is squarefree and $\not\equiv 2\bmod 4$
 
@DickSquizer Write out the definition of integral closure.
 
@PedroTamaroff every element is integral over $\mathbb{Z}$, meaning it satisfies a monic polynomial
 
11:04 PM
@DickSquizer Aha. So, you want the smallest subring of $\Bbb Q[\sqrt{n}]$ that contains $\Bbb Z$ such that...?
 
such that given $\alpha\in\mathbb{Q}[\sqrt{n}]$, $\text{min}_\alpha\in \mathbb{Z}[x]$ is monic
wait do you mean smallest or largest
 
And what have you tried? For example, is it clear that if $R$ is the closure, $\Bbb Z[\sqrt n]\subseteq R$? The hard part is to show $R$ is inside $\Bbb Z[\sqrt n]$.
 
I don't see why $\mathbb{Z}[\sqrt{n}]\subseteq R$.
 
Wait, I told you nonsense.
Sorry.
 
@r9m When I started to work on it I had no idea about the way the closed form looks like.
 
11:08 PM
The integral closure are the elements in $\Bbb Q[\sqrt n]$ that are integral over $\Bbb Z$.
@DickSquizer Sorry about that.
 
That isn't different from what I said before is it?
before "wait do you mean..."
 
I was wrong, you weren't.
 
Okay. Just making sure.
 
So, it is trivial to check every element of $\Bbb Z[\sqrt n]$ is integral over $\Bbb Z$.
 
r9m
@Chris'ssis I'm not sure how I even begin to attack the problem ..
 
11:10 PM
i don't see why.
 
If $\alpha=a+b\sqrt n$, what is a monic polynomial with integer coefficients that kills it?
Here $a,b\in \Bbb Z$.
 
@PedroTamaroff yeah that's the part i don't see.
 
Well, can you find any polynomial that kills it?
 
$x-a-b\sqrt{n}$
 
Well, that hasn't integer coefficients.
$(\alpha-a)^2$ is equal to?
 
11:12 PM
$b^2n$
ah i see wha toyu're getting at
 
Well, then $(X-a)^2-nb^2$ is monic and kills your number.
 
$x^2-2ax-a^2-b^2n$
 
You're missing an $X$.
$X^2-2aX+a^2-nb^2$.
 
yes, thanks
alright well that's one half so let's see
we need to show that if $\frac{a}{b}+\frac{c}{d}\sqrt{n}$ is integral then $b=d=1$
 
Well, the other part is to show that if $q+r\sqrt n$ is integral over $\Bbb Z$, then $q,r$ are integers.
 
11:15 PM
well i can show that $(x-q)^2-r^2n$ is in $\mathbb{Z}[x]$ if and only if $q,r$ are integers
But that doesn't help me because we need an arbitrary monic
 
Right.
However....
$\sqrt n$ has the good property that its square is an integer.
So...?
The claim here would be that we can assume the monic polynomial is of degree two.
Clearly it cannot be of degree $1$.
 
don't see why
 
What don't you see?
 
$(q+r\sqrt{n})^2=q^2+r^2+2qr\sqrt{n}$ alright, so we need to subtract off a $2qr\sqrt{n}$ which means we need $2q\left(q+r\sqrt{n}\right)$
so x^2-2qx-3q^2-r^2
 
@r9m I had the same problem at the beginning. It's a deadly integral.
 
11:21 PM
does that work?
 
I'm out. I need to take some sleep.
 
@DickSquizer The simplest guess is $(X-q)^2-nr^2$. =)
At any rate, my claim is the following: if $q+r\sqrt n$ is integral, then we can find a quadratic polynomial that kills it.
Can you prove that?
 
i just don't get the chain of logic here. that's a good simple guess but we don't know that it has coefficients in $\mathbb{Z}$.
 
Of course, of course.
 
"integral"
 
11:23 PM
One needs to work a bit more.
 
i can't prove it for a quadratic poly in $\mathbb{Z}[x]$
 
Boo @Pedro @Mike
 
Well, suppose that $X^n+a_1X^{n-1}+\cdots+a_{n-1}X+a_n$ kills $q+r\sqrt n$.
Now write that as $(X-q+q)^n+a_1(X-q+q)^{n-1}+\cdots+a_{n-1}(X-q+q)+a_n$
 
Upon expansion and multiplication by the denominator of $q$ to a suitable power, we get what?
 
11:25 PM
honestly i don't see what you're getting at
 
We have $(X-q)^2=nr^2$ now!
 
Why are you insane, @Mike?
 
Hi, @Ted!
 
@KarlKronenfeld Am I being clear here?
 
I just don't see how that helps
 
11:26 PM
Hi @Karl :)
 
@Ted It's garbage! It has garbage answers! Especially the one with 62votes.
 
@DickSquizer Well, consider the case that $n=3$.
For the sake of argument.
 
alright so we're looking at $(x-q+q)^2+a_1(x-q+q)+a_0$
 
We know that $${X^3} + {a_2}{X^2} + {a_1}X + {a_0} = 0$$ kills our number $r+q\sqrt n$.
@DickSquizer Sorry, we want to reduce the polynomial to a quadratic one.
I am not sure if this will work, though. The polynomial might not end up being monic. Let's see.
 
This place is full of garbage, @Mike.
 
11:29 PM
@Ted Usually the garbage isn't so highly upvoted
 
@KarlKronenfeld Do you think my idea will help?
 
@PedroTamaroff I can't tell what the question is.
 
okay i believe it
 
@KarlKronenfeld Proving the integral closure of $\Bbb Z$ over $\Bbb Q[\sqrt n]$ is $\Bbb Z[\sqrt n]$.
 
wait no i don't
wait yes i do
 
11:32 PM
@Mike: Don't lose sleep over it.
 
Let's see if Karl can help.
 
@Ted It turns out I'm not teaching algebra at all.
 
I don't know anything about all this. =P
 
What are you doing? @Mike
 
$(x-q+q)^2=(x-q)^2+\text{ other terms of degree at most one }+q^2=nr^2+q^2+\text{ other terms of degree at most one }$
 
11:33 PM
So it looks like one containment is obvious, but the other one is the one you guys are working on, right @PedroTamaroff
 
@KarlKronenfeld Ayup.
@DickSquizer You said $n\not\equiv 2\mod 4$, also.
 
I really wish people would stop using the word obvious in this discipline.
 
@DickSquizer Won't happen.
=)
 
I don't even think about it, unfortunately.
 
11:34 PM
Obviously!
 
@PedroTamaroff it's also squarefree
but i don't know if those two things are relevant
 
@DickSquizer Yes, sure.
 
@Ted Pre-algebra... expanding 3(x+4)
 
@Mike WAT.
You're dealing with math unborns. =O
 
@Pedro That was hard for you once too.
 
11:36 PM
@Mike Hey, I'm not criticizing. But that looks like school, not college.
 
They're people trying to learn life skills for real, so it's important.
He's teaching adults, @Pedro.
 
@Pedro I'm not teaching a college class.
 
It's the beginning of a problem. I'm still trying to understand the first sentence, which is stated "Let $n$ be a squarefree number not congruent to $2$ mod $4$, thus the integral closure of $\mathbb{Z}$ in $\mathbb{Q}[\sqrt{n}]$ is $\mathbb{Z}[\sqrt{n}]$"
haven't even gotten to something they've asked me to prove yet
 
I still say tax forms is a good place to tackle percentages and basic algebra, @Mike.
 
@DickSquizer Do you have any lemmas or tools to work out integral closures?
@Mike What class is that?
 
11:38 PM
@Ted Not my choice... my job is to help them while they work on worksheets.
 
@DickSquizer That ought to be "not congruent to $1$ mod $4$".
Heya, @TedShifrin
 
Seems @Pedro is ingoring me today!
 
@DanielFischer yes, sorry
 
Hi @DanielF
 
@DanielFischer okay so apparently you get why then
 
11:39 PM
@DickSquizer You need to do something about your username.
=D
 
Easier for you, then, @Mike. Just be calm and patient.
 
@PedroTamaroff why?
 
@TedShifrin Why so? =O
@DickSquizer "Dick Squeezer...?"
 
@DickSquizer You know already that all elements of $\mathbb{Z}[\sqrt{n}]$ are integral?
 
@DanielFischer We proved that, yes.
 
11:40 PM
@PedroTamaroff Heh, yeah, been gettin that one a long time
I tried going by Richard but it just gets back to this eventually. May as well not fight it.
 
I hope I haven't offended you now...
 
Nope, my skin is thick.
 
@Ted Absolutely. The student I spent most of my time with usually did fine - it was a matter of softly pointin out where they went wrong, and congratulating them when they don't.
 
@DickSquizer Let's hear Daniel out, then. @DanielFischer
 
@Pedro There's nothing wrong with squeezing dicks.
 
11:41 PM
@DanielFischer yes we do
We are working on the reverse containment
 
Well, maybe I was just being stoopid, and one should use the integral root test or something on $(X-r)^2-nq^2=0$.
 
@DickSquizer Okay, so let $z = \alpha + \beta\sqrt{n}$ be integral. The case $\beta = 0$ is left as an exercise. For $\beta\neq 0$, we know that $(X-\alpha)^2 -\beta^2\cdot n$ is the minimal polynomial of $z$ over $\mathbb{Q}$.
 
@DanielFischer Agreed.
 
Now, $z$ is integral if and only if its minimal polynomial lies in $\mathbb{Z}[X]$.
 
@DanielFischer why?
 
11:44 PM
Monic?
 
That's the definition!
 
One Gummibear to @Ted.
 
@Ted Minimal polynomials are monic by def'n.
 
yeah but only if.
maybe the minimal polynomial is not monic, then $z$ is not integral
 
Over a field, yes.
 
11:45 PM
Oops... I see now that you realize that.
@TedShifrin Oh, shush.
 
@DickSquizer If some other polynomial kills $z$, the minimal polynomial divides it.
 
Yeppa.
 
@PedroTamaroff but what if none of the polynomials that kill $z$ are monic?
 
And if the other polynomial is monic with integer coefficients, Gauß' lemma says what?
 
@DickSquizer The minimal polynomial always kills it, the thing is what happens with its coefficients. Unless the element is trascendental.
 
11:47 PM
@DickSquizer Integral means that it is the zero of a monic polynomial with integer coefficients.
 
<--- still not sure what to do with gummibear
 
@TedShifrin You eat it?
 
okay so we're letting $z$ be integral over $\mathbb{Z}$ which means it satisfies a monic polynomial in $\mathbb{Z}$. so its minimal polynomial is in $\mathbb{Z}[x]$.
 
Just what I need. More calories making me fat.
 
i don't know if i believe the only if but i think that isn't relevant to the direction of this proof
 
11:50 PM
@Ted Give up, eat the gummibear.
 
so... what does this get us
 
@DickSquizer You explicitly know the minimal polynomial, and can from that find out that it has integer coefficients if and only if ...
 
@DanielFischer we know the minimal polynomial over $\mathbb{Q}$
 
but how does that get us the minimal polynomial over $\mathbb{Z}$
 
11:52 PM
$$X^2 - 2\alpha X + (\alpha^2 - n\beta^2) \in \mathbb{Z}[X] \iff 2\alpha \in \mathbb{Z} \land (\alpha^2 - n\beta^2)\in\mathbb{Z}.$$
10 mins ago, by Daniel Fischer
Now, $z$ is integral if and only if its minimal polynomial lies in $\mathbb{Z}[X]$.
 
actually i don't see why the minimal polynomial has to be in $\mathbb{Z}[x]$. the minimal polynomial of the thing over $\mathbb{Z}$ is what we know divides the monic polynomial, why can we relate that to the known minimal polynomial over $\mathbb{Q}$
 
@Mike Holy Shit.
Is Jack Lee the author?
 
Yep.
 
@DickSquizer $z$ is integral if and only if its minimal polynomial over $\mathbb{Q}$ lies in $\mathbb{Z}[X]$.
 
@DanielFischer i know that's what you're saying but i still don't see why
 
11:58 PM
@DickSquizer The one direction is clear, if the minimal polynomial is in $\mathbb{Z}[X]$, the number is integral.
So the "only if".
 
@Pedro @Mike: I've known Jack since he took a grad course from me at MIT.
I posted a comment.
 
Let $m_z$ be the minimal polynomial of $z$, and $f\in \mathbb{Z}[X]$ monic with $f(z) = 0$.
 
@TedShifrin Cool.
 
@DickSquizer Then we know that $m_z \mid f$ in $\mathbb{Q}[X]$, right?
 

« first day (1357 days earlier)      last day (3664 days later) »