« first day (637 days earlier)      last day (4683 days later) » 

06:07
Morning.
Yay : )
Morning ppl.
Hello Gigili.
Oh @rob, now that you stand as a moderator, I will have no chance.
@Gigili I am not a moderator yet. I am just running to be one. And why would that be a problem?
@robjohn Awesome! You have my vote. (assuming I can vote)
06:10
@MattN Assuming?
Don't know, maybe it's the current mods who select the new ones.
@MattN Sure you can, you have the rep. Now all you need is to be there when the polls open.
Huzzah!
@MattN Nope, we vote for mods.
And the other one I want is mixedmath.
@JM ?
06:11
@MattN I mean, current mods can vote for their favorite candidates, but not unilaterally. :)
@JM Ok : )
@robjohn I know and it's no problem. I'm just saying I'll have no chance if I stand as well.
Which means you'll knock everyone out. Good luck.
"I type softly, but carry a big button!!"
@robjohn what did you use for $x$ to evaluate the $A_k$s? Or, where did you get an $a_4$ in the denom of the fraction?
@Gigili there are two mods being selected, so even if what you say is true, which I doubt, there is another spot.
06:14
"Complicated J.M."
@Gigili How so?
@Jeff I scaled the previous answer :-)
@JM I don't get what you mean 99.99% of the time!
@Gigili That's okay. I don't get myself half of the time, also. ;)
i have no idea what that means. I tried using $x=-\omega^k$, that makes the numerators $0$. I tried using $x=-a^4\omega^k$, because the $a^4$ was in your correct answer, but then the numerators evaluate to $1-a$ instead of $1-1$.
06:17
bbl
@Benjamin: aha, I figured it out
@Jeff For $\frac{1}{x^5+a^5}=\frac{1}{a^5}\frac{1}{(x/a)^5+1}$
it took me a bit
@Jeff Then use the previous result with $(x/a)$ in place of $x$
oh. well that's easy. :D
06:20
This would be relevant in parsing JM's allusion.
Man, Roosevelt isn't so well-known nowadays, I see. I'm getting old...
@JM you think you feel old.
Don't feel bad. In my last post, I did the same...
On that note: probably a good thing that the nomination texts are relatively low key...
@mixedmath Ok
06:25
@Benjamin: So I was trying to think of the right characterization of absolutely flat things to use here, and I want to use that absolutely flat local rings are fields
yes
Let's have a quick check to see why this is the case
suppose $R$ is not a field
then take $x \in R$ such that $x$ is not a unit
then look at $(x)$
Oh I forgot to say that $x \neq 0$
then because principal ideals are idempotent, a quick corollary of this is that $(x) = (ax)$ where $a$ is some element in the ring such that $x = ax^2$
Furthermore $ax$ is idempotent
@mixedmath I would just multiply both sides by $1+\cos(\theta)$ and everything falls togehter.
but then $R$ is local so that $ax = 1$ or $ax = 0$
Now we cannot have the second case for then this would contradict $x \neq 0$ because:
$x = ax^2 = x(ax) = x(0) = 0$
@rob: the other two answers did that too. In hindsight, I'm surprised I didn't see that first. But... I didn't. I was really just going for another said and lost Roosevelt analogy
so we must have $ax = 1$ contradicting $x$ being not a unit
06:28
great
so that R must be a field
Ok
so start with some prime ideal p and a maximal ideal m containing p, all in $A$
It's funny how $\sec$ and $\csc$ are virtually unheard of here in Denmark while apparently they show up again and again in US education. They are pretty much never discussed in highshool(-equivalent...ish) maths
(if I recall, $A$ is your ring, $R$ its nilradical)
06:30
yes
then localize $(A/R)$ at m
so it's a field
@kahen I think the Americans are just very obsessed with having the complete set of six...
um
how did you....??
@JM but it's not complete. They forgot vercosin, hacoversin, havercosin, etc.
well, we know $A/R$ is absolutely flat, right?
06:32
We used to have HUGE tables and lots of formulas with those. Now we have calculators, so why bother?
ah pffff
@kahen Oh, I think the versed sine is a handy thing myself for numerics... but it just didn't catch on.
$(A/R)_m$ is a local ring
sorry man
my bad
ah, got you
it happens
forgot that a maximal ideal is prime :D :D :D
06:33
I do that all the time, still
anyhow, this is handy, because we know how many prime ideals are in a field
haha perhaps excusable for a second year undergrad then :D :D
only $(0)$
but we also have a great correspondence between prime ideals of $(A/R)_m$ and prime ideals contained in $m$
peeks in
is it safe? (marathon man reference)
welcome to the party, David
@DavidWheeler There is no non-normal non-separable extension of degree 5. I proved it.
06:36
good on you
good on ya'!!!
@mixedmath Wait
we don't know if the map $f: A/R \rightarrow (A/R)_m$ is surjective
did you write it up somewhere?
@DavidWheeler no but the proof is really easy
shoot
i love an easy proof...don't have to wine and dine them so much...
@mix: consider yourself lucky that you do not go by your real name here...
06:37
oh, another blue David?
no kidding - I'd get so lost in the confusion
yes indeed
@mixedmath How can you infer the ideal correspondence?
it's a standard result that prime ideals in a localization are in bijection with prime ideals contained in the localizing prime ideal
what text are you using for the course?
@mixedmath I am blind
Both Atiyah and Matsumura use this a lot, but maybe not Eisunbud
06:39
the prime ideals of the localisation
oh - no problem
can you finish the argument from there?
are in one to one correspondence with the prime ideals in $A$ that don't meet $S$
it's at the edge
I am so blind !!!!
baaaah
@mixedmath Commutative algebra is self study now
Ah I am so blind.......
is a localization a partial ring of fractions?
06:41
self-study - nice
and the prime ideals in $A$ that don't meet $S = A - m$ are precisely those that are contained in $m$.......
a partial ring of fractions... I suppose so
i'm thinking of something like $S^{-1}R$
@mixedmath
Ok so since our prime ideal $p \subseteq m$
06:44
we have that it's extension in $(A/R)_m$
say we call it $p_m$
is a prime ideal in $(A/R)_m$
ohohohohohohoh
in general, we don't know that an extension of a prime is prime
@mixedmath Sorry I said something wrongly
@Jeff yes?
but basically I want to say something like the image of $p$ in $(A/R)_m$ is zero something like that to infer that $ p = m$
@robjohn what i posted was wrong.
the latex was wrong
@robjohn $\displaystyle\frac{1}{a^5}\frac{1}{(\frac xa)^5+1} =\frac{1}{5a^5}\left(\frac{1}{(\frac xa)+1}+\frac{\omega^1}{(\frac xa)+\omega^1}+\frac{\omega^2}{(\frac xa)+\omega^2}+\frac{\omega^3}{(\frac xa)+\omega^3}+\frac{\omega^4}{(\frac xa)+\omega^4}\right)$
06:49
let's take a step back and reconsider what we know. There is one prime ideal in $(A/R)_m$, which corresponds to a unique prime ideal in $A/R$ contained in $m$
but this then corresponds to a unique prime ideal between $R$ and $m$
@rob another $a$ factors out on top, making the leading fraction $\frac 1{5a^4}$... whew! :D
now I think I'm ready to go on to the integral.
@mixedmath I think you mean it corresponds to a prime ideal between 0 and $m$
because recall the prime ideals in the localisation are in one to one correspondence with the prime ideals of $A/R$ that don't meet $A - m$
that is the prime ideals that are contained in $m$
that's true, but it's also true that prime ideals in $A/R$ contained in $m$ bijectively correspond to prime ideals in $m$ containing $R$
by unrapping what it means to quotient a ring
06:53
you mean in $A$ containing $R$
yes that's the correspondence theorem
I do mean in $A$, but in particular in $m$ in $A$
and every prime contains $R$, as it's the nilradical
so both $p$ and $m$ contain $R$, and are both prime
so...
06:54
$m = p$
exactly
and that's that
wait
I am trying to understand your ideal correspondence thingy
me too, but i am slower than dear ol' Benjy
you're saying if we have a maximal ideal $m$ in $A/R$
then the prime ideals contained in here
are in bijection with those in $A$ containing $R$ and contained in $\pi^{-1}(m)$?
$\pi : A \rightarrow A/R$
06:57
we have bijections between:

(primes in $(A/R)_m$) - (primes in $(A/R)$ contained in $m$) - (primes in $A$, contained in $m$, containing $R$)
yeah it's the bijection on the right that I have not seen.
oh, perhaps the confusion is in $m$ and its notation
so, recall that we defined $m$ in $A$ originally
so when we refer to $m$ in $(A/R)$, we refer to the image of $m$ in the quotient
So this is like a double correspondence theorem
or rather, we're using two different correspondences; yes
So let me run through your argument again.
Let $p$ be a prime ideal in $A$
Then there is a maximal ideal $m$ that contains it
Now we know that $(A/R)_\bar{m}$ is a field
where $\bar{m}$ is the image of $m$ in the quotient
07:02
Anyone can help me? I'm trying to solve $x^3-y^3=xy+61$ which is $(x-y)(x^2+xy+y^2) = xy+61$ And then I'm stuck
@mixedmath I got it
great!
the only prime ideal in $(A/R)_\bar{m}$ is $0$
maybe not....
no, you're right so far. The only prime ideal is (0) there
I want to say something like the preimage of $0$ in $A/R$ is something @mixedmath
07:07
as in, you're trying to stay away from the first correspondence? it's a bit unclear what the preimage actually is apriori
@Gigili Are you looking for a solution in integers?
@mixedmath Ok like this:
in the original ring $A$ we have $R \subseteq p \subseteq m$
so since the map onto the quotient is surjective, we have that
$f(p)$ is a prime ideal contained in $f(m)$
Yes @BrianMScott. I asked on the main site as I didn't get any response here!
since we are localising by $f(m)$
$f : A \rightarrow A/R$
Now $f(p)$ is in one to one correspondence with the only prime ideal of $(A/R)_\bar{m}$ which is zero
07:09
you really like these surjective maps
@Gigili You should add that stipulation to your question on the main site. (I don't know whether I'll have any ideas, but I'll give it a few minutes' thought, at least.)
@BrianMScott I will add it now, thank you.
@mixedmath hmmm ok now I am dead in the water
so right now, we know that since there is exactly 1 prime ideal in $(A/R)_\bar{m}$, there is exactly 1 prime ideal in $(A/R)$ contained in $\bar{m}$
yeah
that is our $f(p)$
Ah
so that there is exactly one prime ideal contained between $m$ and $R$
forcing $ p =m$
got it got it got it got it got it got it!!!!!!!!!
07:13
great!
@mixedmath Now one thing left to understand
is it true
that any $S^{-1}A$ module is of the form $S^{-1}M$
for some $A$ - module $M$
@mixedmath Because I need that result to prove that if $A$ is absolutely flat then so is $S^{-1}M$
@BenjaminLim $\huge{!!!}$. :-)
3
@BrianMScott I have spent the last 3 hours trying to understand this problem....
07:16
I know how you feel.
@BrianMScott Isn't it possible to solve the equation?
@BrianMScott I think at any level even if I were to be in third year I would find this material just as hard
@Gigili I don't know; it looks like the sort of thing that either requires a clever trick or is just plain hard.
Aha, okay.
hmm, it's not immediately apparent to me either way
07:19
yeah........
because
we used the result in our proof
that
if you $A$ that is absolutely flat, then $S^{-1}(A/R)$ is absolutely flat
and we can quickly prove it like this:
strictly speaking, I only used that if $A$ is absolutely flat, then $A_m$ is a field for any maximal ideal $m$ - perhaps you're thinking to prove that result?
yeah
oh but wait
If you switch tabs fast enough with sufficiently unfocused vision, you read "money commutes with everything" on the star panel.
@mixedmath Did you not use the fact that $A_m$ is tacitly flat ??
no, I don't think so
where would I have done that?
07:28
Because don't you want to say that $A_m$ being a local ring + absolutely flat makes it a field?
if I might summarize my argument, I used that $A/R$ is abs flat means $(A/R)_m$ is a field
Would the set of intervals $(C-n,C+n)$ (where C=the cantor set) be a countable union of nowhere dense sets?
@mixedmath hmmmm
oh, perhaps I justified that result. That's how I know how to prove that $(A/R)_m$ is a field
I don't know what you do or don't know, and I don't have my copy of Atiyah here to see
hmm
ok, could you tell me what is your way to prove that $(A/R)_m$ is a field
07:31
you were right earlier - $A_m$ local and abs flat makes it a field
and it
it's true that $S^{-1}A$ is abs flat if $A$ is abs flat
the only way I can think to prove that right off is very functorial
Please no Tor or Yoneda Lemma
I know zero homological algebra :D :D :D
Anyway this is good enough for me for the moment
a lot to take in :D :D
@mixedmath Learning AM by one's self is crazy man
07:33
A good exercise, that's for sure
but it's crazy :D :D :D
madness
anyhow, congratulations, and good night - I'm off to bed myself
night man!!
thanks so much :D @mixedmath
@anon I read "money doesn't grow on trees", how fast and unfocused it needs to be?
07:37
Sure thing. I am trying to be nice even to my enemy.
@Gigili I see no "doesn't," "grow," or "trees," so perhaps you were ctrl-tabbing at high speeds.
Might be it, I'm so stressed out due to moderator elections!
A rule which is not too silly is: have as few enemies as possible.
A corollary to that is: don't spend the few spots for enemies on has available on math.SE... it is simply a waste! There are so many people out there who __*REALLY*__ deserve enemyhood!
A deep philosophical point, I need time to get it!
IOW, right beside the classical «Thou shalt not covet thy neighbour's house,» I'd put «Thou shalt find your enemies among your neighbors, not random people out there on the internet!»
07:43
Why not? There is enough space for all of them!
because it is a bit silly :/
it's generally a manifestation of the already classical someone-is-wrong-on-the-internet phenomenon xkcd.com/386
today's xkcd thingie is pretty pertinent to us, btw :)
@BenjaminLim, why do you want to delete math.stackexchange.com/questions/136617/… ?
do we have a quick guide to tex?
08:05
@MarianoSuárezAlvarez that's one of the funniest comics in history!
indeed
we all recognize ourselves in it
xkcd has its moments, but I don't think it's very consistently funny. I prefer SMBC.
your avatar changed :D
After much public consternation!
For a good amount of time, it would show my previous grav in some conditions and my new one in others. Conditions were things like: who you were, which browser you used, and what ratings or size settings you had for viewing gravatars. It was rather incoherent.
@anon never heard of SMBC, but I agree about xkcd's inconsistency (and the same can be said about PhD comics)
08:12
@Jeff The odds are pretty much 50-50 whether Randall would be funny or straightforward...
whois Randall?
the guy who does xkcd, randall munroe
@anon oh. but when he hits, he's on (like that internet one).
@Jeff True. One has to gamble when making people laugh, and inevitably you'll miss sometimes. But jackpots aren't impossible.
@JM i'm still waiting for mine :D
say, can anyone in here explain laurent series and how to use them to do an integration to me in 6 hours? :D
08:23
The tricky bit for Laurent series is if the poles are within the integration interval...
they are within the integration interval (I think). i have five (very similar) integrals. the first is $\int_0^\infty \frac{1}{x+a}\ dx,\quad(a > 0 \in \mathbb{R})$.
@Jeff you can save a bit of typing by doing \mathbb R instead of \mathbb{R}: $\mathbb R$.
@kahen ty
And similarly for \frac: \frac12 instead of \frac{1}{2}: $\frac12$.
You can save even more just typing \Bbb R
08:28
I love all the laziness tips... :D
Use of \Bbb is discouraged these days I think
Oh? Why?
@JM it's not just laziness. It makes one's LaTeX source easier to read when writing
I think more often than not curly brackets actually make reading others' source easier, personally.
«these days» = «with all likelyhood, since before you entered highschool», actually :)
08:30
@kahen That's what I meant too. If you write things nicely, then it takes less effort to read them when you go through them again. :)
@Mariano: So, before 2005? :P
:D
testing... $\Bbb R$
Of course the real smart ones do things like this: \usepackage{xspace} \def\RR{\ensuremath{\mathbb R}\xspace} ... As \RR is locally compact... Take an $x \in \RR$... :)
Or you could fiddle with unicode and unicode-math.sty to just let you type in the glyph directly in your TeX source :)
ok, back on topic: can someone just tell me how to do that integral. i hate to just ask questions, but i've been working on this problem for 14 hours straight (and 1 week) and need to get it done soon (like very soon - oh, I can't get Maple to do it, either). $displaystyle\int_0^\infty \frac{1}{x+a}\ dx$?
08:35
@anon, at least before 1997
That was before I even liked math!
Write $u=x+a$.
$\unicode{x211D}$ hmm, it seems to work nicely.
ok. you think this is just a regular Reals integral? My instructions are to use the residue theorem (this is just one of the five which the original integral broke into usingpartial fractions).
08:39
@Jeff You serious?
You can prove the divergence even without using complex machinery...
yes (about what? but 'yes', whichever part you're asking about)
Hey, you asked one simple integral, I gave one simple answer. I didn't realise that the question was something completely different.
@DavidWallace well, i thought i needed to factor it to get to use the residue theorem. did i go about it the wrong way?
Oh, the right answer to the wrong question. A common problem, that.
08:41
@davidwheeler actually, based on what i'm reading now, it looks like i didn't need the partial fractions. the residue theorem just relies on the singularities
What's "it", @Jeff? You haven't told me the real question!
@JM it's especially common among the answers I receive :D
Aha, somebody pinging Wheeler instead of me for a change. Usually it's the other way around.
the original question is $\int_0^\infty \frac{1}{x^+a^5}\ dx$, where $a$ is real, positive.
Davids... so many Davids...
08:43
so i guess i have the singularities, they are $e^\frac{ \pi i}{5}, e^\frac{3 \pi i}{5}, e^\frac{5 \pi i}{5}, e^\frac{7 \pi i}{5}, e^\frac{9 \pi i}{5}$
So how does the 5 make any difference? You can still write $u=x+a^5$ and it becomes very easy.
Hmm, we have a New-Zealand PhD student and he was showing me this "haka". What the!
Apr 11 at 6:07, by anon
Damn you Davids and your nefarious first name.
@DavidWallace: Typo, it's $x^5+a^5$
@DavidWallace typo: $x^5+a^5$ in the denominator. $\displaystyle\int_0^\infty \frac{1}{x^5+a^5}dx$
And the singularities don't make too much difference if you're going from 0 to infinity, because there'll usually be a contour that avoids them.
08:44
@anon Just to be above board: your real name isn't David?
It is not. :-)
I still have nightmares about Davids.
My real name is David. What on earth is wrong with the haka?
@DavidWallace but the countours still go around them, right?
I would surely make a good fit.
@DavidWallace Nothing, it just looks funny 8-).
08:45
@JonasTeuwen You Jonases present a few problems, too... ;)
@JonasTeuwen My god, say that out loud, and you're likely to get beaten up!
@JM Hmm, yes...
@Gigili whereas my dreams about Gigilis are nothing but sweet.
@DavidWallace Well, I actually laughed, and I didn't get beaten up! :D.
@JonasTeuwen You mean he performed it himself, or he showed you a video?
08:47
Both!
The video was funnier.
With the tongue and stuff...
@Jeff So, yes, break it up into partial fractions. Try to make each one something that you can integrate individually.
@JonasTeuwen Piri Weepu is nothing but scary.
Yeah! It was that guy!
@DavidWallace is there a way to use residue theorem without doing partial fractions? i've already done the partial fractions, but i guess i didn't have to.
@DavidWallace is there a way to use residue theorem without doing partial fractions? i've already done the partial fractions, but i guess i didn't have to.
@DavidWallace I just sent you a blah, see the blah.
did my last post come up? I keep getting a "retry/cancel" message
08:53
Yes, twice.
To use the residue theorem, you'd need to do some kind of transformation that turns 0 to infinity into some closed contour. It's not obvious to me how to do that.
@DavidWallace oh. so then i should use the partial fraction decomp. i'll get five Ln functions.
That's how I'd do it, sure. But is it homework where you HAVE to use the residue theorem?
@DavidWallace Yeah. Haha!
Then I don't know how to do it, sorry. Kind of out of practice.

« first day (637 days earlier)      last day (4683 days later) »