« first day (1856 days earlier)      last day (3164 days later) » 

2:00 PM
@TheArtist You said it as if I knew yours. ;) Chris's sister the artist isn't good enough? :D
 
@DanielFischer I seem to recall Hardy providing some sufficient conditions, but of course I can't find the paper or book at the moment. The conditions I stated are most likely a misinterpretation of what he stated.
 
@Chris'ssistheartist no but I am trying to quote you :P
 
@Chris'ssistheartist fine, that will do .
 
@RandomVariable I never asked you if you attend integration for many years. No idea about sos either (excepting the period in which I joined MSE - he was already in place with great answers).
 
2:02 PM
@RandomVariable Possible. But I have no idea where you should look.
 
@Chris'ssistheartist A daring journey on a $\int$
 
@Chris'ssistheartist I didn't even take a calculus course until college.
 
@Chris'ssistheartist google told me that "escapading" is not the verb of "escapades" , so it does not really work.
 
@TheArtist It's not easy to find a good name for a book. :-)
@RandomVariable I see.
 
@Chris'ssistheartist really?, making a book is easier than choosing a proper name ?
 
2:07 PM
@Agawa001 I only said it's not easy to find a good name for a book. :-)
 
@Chris'ssistheartist totally agree with you.
 
i wonder how much not easy to write a book that the name isnt even easy to find
 
@Chris'ssistheartist would not say its easier than writing a book. Writing a book is very difficult.
 
@TheArtist It's very difficult but at the same time it's an amazing journey, you learn so much while doing that. :-)
 
if i made a book, i would find a suitable name for it
i would name it for example, "what you dont expect to read in other books" lol
2
 
2:11 PM
:-)))))))))))
 
@Chris'ssistheartist This came to my idea :P Title : "Finding True Love" . Centered below the text is the sign \int ♥ dx
 
@TheArtist :-)))))))))) People would think it's about some novel. ;)
 
@Chris'ssistheartist yes correct. And its important to have what the book is about in the title
@Chris'ssistheartist : "Living Inside an Integral"
@Chris'ssistheartist sharing treasures and things you found in life by living with integrals ^^^ ;)
 
@TheArtist lollllllll, nooooooooooo :-)))))))))))
Back in 30-60 min.
 
@Chris'ssistheartist lol
 
2:17 PM
working offline is time saving, take this advice from a programmer :)
see ya
 
@Agawa001 language?
 
maternal language is C
i know java and javascript and vb
as well
and you cant be a complete programmer if you dont know some asm basics
so i do
 
@Agawa001 is that your profession too? :)
 
i guess so, i spend lot of time programming stuff
 
@Agawa001 cool ^_^
 
2:25 PM
its really diificult, sometimes you cant satisfy people's needs, sometimes you repeat something done dozen times (deja vu), so it turns annoying
but the main thing is fun in its globality
 
@Agawa001 it must be. I failed my "Introduction to Programming for engineers" course
@Agawa001 and we did C and Matlab.
 
lol, have you tried harder ?
 
@Agawa001 I have done HTML , super easy lol
and CSS
 
there s nothing easier than programming, its like you tell the machine to do something, and it does nothing apart what you order it
:)
 
@Agawa001 i put that aside now. Will have to repeat it next year so will try more then, but yes I did word hard.
 
2:28 PM
just take it as simple
 
I think its an easy subject for some people and not so good for other people.
 
The sum of roots of the equation $z^{2014}=i$ is equal to what? How can we know?
 
its super easy to understand a dumb-machine, what i find harder is understanding social life :(
 
@user159870 De Movire's?
 
@user159870 Generally, when you have a polynomial, how is the sum of the zeros connected to the coefficients?
 
2:30 PM
@Agawa001 what is so hard about social life?
 
@M.S.E Somehow, people tend to be involved. And people are difficult.
 
@DanielFischer involved in?
 
@M.S.E well, not all people think the way you do, and thats lil bit conflictual
 
@M.S.E In social life.
 
@Agawa001 that is why you talk to as many people and find the ones that share common interest and take it to the next level to build a relationship.
@Agawa001 Isn't it a game of probability? The rest (majority) of the people you say "see you around" and never talk again.
 
2:34 PM
@DanielFischer You mean this? the sum of the roots is $\frac{-a_{n-1} }{a_n}$, $a_n$ is the leading coefficient and $a_0$ the constant.
 
@user159870 Yes.
Now, what are your coefficients when you want the solutions of $z^n = a$?
 
@M.S.E its complicated, more events can make things more complicated, not just one or two variants, its more thorough
 
@DanielFischer but how are they difficult? :) look at this chat, no one is difficult here.
 
@M.S.E You are new here, right?
 
@DanielFischer lol yeah but still so far, in hours of a few weeks I have been here :P
 
2:38 PM
@M.S.E wait to see questions of topology, you would probably change your mind
 
@DanielFischer $a_n=1, a_{n-1}=0$.
Yes?
 
@Agawa001 okay :P
 
if 1+1 =2 why 1-1 ≠ -2 ?
 
@Agawa001 but similar thinking people gives rise to less complication rate like 50% will have no problem.
 
@user159870 Yes. (For $n > 1$.)
 
2:41 PM
How can PI exists if circle aren't provable?
 
@lopata Because + means going right 1 step in the number line, and - means going 1 step left on the number line
 
@M.S.E i can find 70% of people who think in my way just in internet (usually people are more plugged into material)
 
How do you prove PI exists?
 
@Balarka I have to prove that $G \cong Z_p$ where G is an abelian simple group. Now since G is simple the only subgroups it will have will be 1 and G. Similarly for $Z_p$. So if I define the map $\phi$ to be the one which takes $\phi(1) = 1$. I got till this much. How should I proceed after this
 
@DanielFischer So there are no roots. Yes?
 
2:43 PM
@lopata a(sign + or -) b=c , means a is the initial position, sign is the direction to move, b is the number of steps to take, and c is the final landing position
 
lol
 
@lopata $\pi$ is defined to be the ratio of circumference to the diameter
Hello@TedShifrin
 
I mean, how do you prove cricumference exists if perfect circles don't exists?
 
@Agawa001 that is true :)
 
@Ted A question :
How do I prove that a group $G$ of even order is isomorphic to $Z_p$ . What happens if $G$ is of odd order
 
2:46 PM
@Rememberme you have to prove the injonction of abelian group G in subgoup Z
 
@lopata Many undefined terms there :
1) What do you mean by perfect circles?
2) What do you mean when you say circumference exists
@lopata injonction ?
 
@Rememberme take an element in your abelian simple group and look at the subgroup generated by a single element from your group.
recall that a subgroup of an abelian group is normal
 
@user159870 No. Counting multiplicities, every polynomial of degree $n$ has exactly $n$ zeros in $\mathbb{C}$. That the sum of the zeros is $0$ doesn't mean there are no zeros.
 
@Rememberme I meant Injection
In mathematics, an injective function or injection or one-to-one function is a function that preserves distinctness: it never maps distinct elements of its domain to the same element of its codomain. In other words, every element of the function's codomain is the image of at most one element of its domain. The term one-to-one function must not be confused with one-to-one correspondence (aka bijective function), which uniquely maps all elements in both domain and codomain to each other, (see figures). Occasionally, an injective function from X to Y is denoted f: X ↣ Y, using an arrow with a barbed...
 
@DanielFischer It's way too much to post here, but here's a link to the paper if you're interested: play.google.com/books/…
 
2:52 PM
@DanielFischer So can we say that the sum of roots of $z^{2014}=i$ is $\frac{a_{2013}}{a_{2014}}=0$?
 
Well isomorphic means there exists a bijection $\phi$ such that $\phi(xy)=\phi(x)\phi(y)$. Now I have to find a bijection which has this property @lopata
 
@user159870 Yes.
 
I can't see latex
 
copy this in your bookmarklet
 
2:57 PM
damn its too long to fit
 
Does your Group G has + ? has . ? has associative and neutral element? @Rememberme
What's bookmarket?
doesnt work for me
 
your favorites
 
thanks I did but I still can't see latex
 
just drag it there
 
3:02 PM
It is a group so its elements have to be associative and has to have an identity element by definition @lopata I don't get what you mean by has +?
 
has + operation
has ● operation
 
@lopata click on it after you dragged it
 
It has to have an operation for which when elements act according to that operation it is associative.
 
ok
 
Got it Balarka thanks !!
 
3:07 PM
does someone know which way I could define PI without circles?
 
chat.stackexchange.com/transcript/message/23859137#23859137 @Balarka If you are free look at this question . G is a simple group of odd order
 
With congruencies? or something like that?
 
every solution to the differential equation $y'' + y = 0$ is periodic with period $2\pi$
so define it to be the period of a nonzero solution of $y'' + y = 0$
:)
 
Nice one @Mike
 
ok thanks
 
3:08 PM
Morning @Mike
 
morning
 
@lopata $$\pi=3+\textstyle \frac{1}{7+\textstyle \frac{1}{15+\textstyle \frac{1}{1+\textstyle \frac{1}{292+\textstyle \frac{1}{1+\textstyle \frac{1}{1+\textstyle \frac{1}{1+\ddots}}}}}}} $$ although that's a "trollish" way of defining it
 
Well @lopata $\pi = \int^{1}_{-1}\frac{dx}{\sqrt{1-x^2}}$
 
ok , MMh yes I guess but we can't predict the numbers so
 
Which is just arclength of the top part of the unit circle
 
3:11 PM
I ma trying to view this with a latex viewer
 
chat.stackexchange.com/transcript/message/23859137#23859137 @Mike Can you look at this question of mine. Here G is a simple group of odd order
 
Ohh it works now
 
@DanielFischer Do you have time to look at a specific case?
 
@Rememberme: I do not understand the question.
 
@DanielFischer i was blocked for 30 mins. I just thought that guy was a troll.
 
3:15 PM
Okay let me write it again
If G is a simple group of odd order . Is it isomorphic to $Z_p$ for some prime p?
Is it also isomorphic when the group is of even order @Mike
 
@RandomVariable Maybe.
 
@Rememberme Lots of simple group of even order.
E.g., $A_5$.
 
Thanks sooo much @DanielFischer
 
It is probably known by the classification that there are no simple groups of odd order.
 
@lopata have you installed ChatJax?
 
3:19 PM
And the thing you said about odd order groups is Feit-Thompson, I think
 
If you mean abelian, you probably have the tools to prove this yourself.
No, @Balarka, that's not sufficient.
 
@MikeMiller Except the abelian ones.
 
No I made a mistake there by saying abelian . I already proved the abelian case
 
I forgot, what was Feit-Thompson's statement again?
 
Odd-order groups are solvable.
 
3:21 PM
ah, ok.
then definitely that's not sufficient.
 
So how should one go by proving that no simple groups of odd order ?
 
Classification of finite simple groups is hard.
 
@DanielFischer Or is the general formula of the sum $(-1)^{n-1} \frac{a_{n-1}}{a_n}$ ?
 
(I'm assuming there's no direct way to prove it, as Mike indicated. But there might as well be, I am not sure)
 
@user159870 No, it's always $-$. Write it as $$a_n \prod_{k = 1}^n (z - \zeta_k),$$ where the $\zeta_k$ are the zeros (appearing according to their multiplicity).
 
3:26 PM
Ok, thanks :)) @DanielFischer
 
@MikeMiller I want to show that every closed orientable $n$-manifold appears as a branched cover of $S^n$. $n = 2$ is easy : pick a genus $g$ surface so that your $g$ torii are squished together in a line. Flip around an axis. This gives your surface minus $2g$ points a $\Bbb Z_2$ symmetry. Quotient to get a branched cover over $S^2$ branched cover $2g$ points.
I don't know how to do this in general for $n > 2$.
ps : the $2g$ points are the points where your axis hits the surface.
 
@BalarkaSen: I'm sorry, of course Feit-Thompson implies that simple groups of odd order are abelian. I do not believe that there is an easier way.
Do you know what a branched cover means for $n>2$?
 
branched over a codim $2$ surface.
should have mentioned that, sorry.
 
I don't even know how to prove it for $n=3$. Why do you think it should always be possible?
 
Because I heard it is.
 
3:31 PM
@DanielFischer The specific case is $$\frac{d}{da} \, \text{PV} \int_{0}^{\infty} \frac{e^{iax^{2}}}{x^{2} - \pi} \, dx \stackrel{?}{=} \text{PV} \int_{0}^{\infty} \frac{ix^{2}e^{iax^{2}}}{x^{2} - \pi} \, dx $$ The justification would seeminly be nontrivial even if the denominator were $x^{2} + \pi$.
 
@MikeMiller Well, that was the statement about every $3$-manifold being a branched cover of $S^3$ branched at a link.
 
@lopata here's a more "trollish" way of defining $\pi$ $$\pi:=\lim_{n\to\infty}\sqrt{6\displaystyle\frac{\ln\big(\prod_{k=1}^nF_k\big)} {\ln(\operatorname {lcm}(F_1,F_2,...,F_n))}}\ \ \ n\in\bf N$$ where $F_n$ is the nth Fibonacci number
 
But yeah, I don't know how to prove that either.
 
Yes, I'm aware of what the statement is. I still don't know how to prove it.
As far as I can tell your claim is false.
 
"orientable"
 
3:33 PM
So?
Ping me when you can tell me why what you just said is irrelevant.
 
Sorry, I should read the paper carefully before commenting.
 
You don't need more than what's in the abstract.
Well, I guess you need to know what a $\pi$-manifold is. They include $S^n$ for all $n$.
 
But... $\Bbb RP^n$ is not orientable, I thought?
 
What is the first example of a $\Bbb{RP}^n$?
 
$\Bbb {RP}^2$.
 
3:37 PM
No.
 
You want a higher dimensional example? $\Bbb RP^3$.
 
I wanted a lower-dimensional example.
 
oh, $\Bbb RP^1$. yikes, that's orientable.
 
Now found out precisely for which $n$ $\Bbb{RP}^n$ is orientable.
Exclude $n=0$ because that's dumb.
 
@lopata Here's an even more "trollish' way $$\pi:=\frac{426880\sqrt{10005}}{A\big[_3F_2(\frac16,\frac12,\frac56;1,1,;B)-C_3‌​F_2(\frac76,\frac32,\frac{11}6;2,2;B\big]}$$ where $A=13591409$, $B =\frac{-1}{151931373056000}$ and $C=\frac{30285563}{1651969144908540723200}$.
 
3:42 PM
Oh, I am dumb. Should have checked $H_n$ before saying something that silly. $\Bbb RP^n$ is orientable whenever $n$ is odd.
 
Sure. Let $M$ be an oriented manifold and $G$ be a free (properly discontinuous etc) group action on $M$. When is $M/G$ orientable?
 
@RandomVariable That's not easy since the derivative isn't Lebesgue integrable. If we split the integral at $2$, say, in the first part we only have to deal with the pole. Without taking out the paper, I would think that part is fine, since the integrals of the derivative converge nicely as the hole shrinks [should converge locally uniformly]. For the other part, I think you need careful estimates. I think it will work out (provided $a$ stays away from $0$), but I'm not sure.
 
Feel free to assume $M$ compact and $G$ finite if you're uncomfortable with the general case.
 
Hi everyone! Help would be really appreciated!

http://math.stackexchange.com/questions/1419718/wanted-a-purely-algebraic-proof-of-the-frobenius-theorem-on-distributions}
 
If you can only formulate it but not prove it, I won't be too broken up. You'll prove it eventually.
 
3:47 PM
@MikeMiller Hmm. Seems like an interesting question.
I am trying to note that $M \to M/G$ is a covering space and use the transfer sequence.
ok, that doesn't work. transfer sequence is $\Bbb Z_2$ coeff.
I'll think about this and get back to you.
 
@IWantToRemainAnonymous wow much wow such troll epic function thanks
 
@BalarkaSen: Anyway, the point was that article provides counterexamples to your conjecture. $\Bbb{RP}^5$ is apparently not a branched cover of $S^5$. It was proved very recently I think that every PL 4-manifold is a branched cover of $S^4$ with locally flat branching set. The Montesinos something something, somebody told me about this but I don't remember.
Because PL 4-folds are smooth 4-folds this says that smooth 4-manifolds branch over $S^4$ with locally flat branching set. This is pretty weak, so I would bet it's either false or not known that they can smoothly branch over $S^4$ with smoothly embedded branching set.
 
@DanielFischer If the denominator were $x^{4}- \pi$, could you definitely say the equation is true without taking out any paper?
 
@RandomVariable Not definitely. Before saying anything definite, I would have to make sure that the part near the pole behaves like I think. But then the other part would be unproblematic by dominated convergence. Assuming of course that $a$ is real. If $a$ is complex, things are different. For $\operatorname{Im} a > 0$, the part at $\infty$ would still be harmless by dominated convergence,
and for $\operatorname{Im} a < 0$, I expect that neither integral exists even as a principal value integral since the integrand blows up then.
 
4:29 PM
@BalarkaSen: Looks like you were right, provided we change the definition of branched cover; see theorem 1. Here we're branching over subcomplexes, but not submanifolds.
 
that looks like a nice set of notes, hmm
 
@DanielFischer I don't recall ever seeing a question posted on main about justifying the interchange of limiting operations in a particular situation where the typical theorems don't apply. But that's usually the type of question I ask you in chat.
 
@RandomVariable Seems not so many people veer out of the "standard theorem" territory.
 
@MikeMiller That seems like a hard question. I am not sure how to do this.
 
You'll figure it out.
 
4:34 PM
@MikeMiller I'm looking at the paper, wait a second.
Um, what's the black and white chessboard coloration on $\beta T$? Monochromatically color it in a way such that if you take a white simplex then the simplices sharing faces with it would be black and vice versa?
 
I didn't read it. You're the one interested in this.
 
4:50 PM
@DanielFischer If the script doesn't detect serial upvoting, is it best to report it or to just let things be? I had 7 or 8 answers upvoted within a few minutes.
 
Hello!!

I want to prove the following lemma:

Let $F[t,t^{-1}]$ be the ring of the polynomials in $t$ and $t^{-1}$ with coefficients in the field $F$ and assume that the characteristic of $F$ is zero.
Then for any $n$ in $F[t, t^{-1}]$, $n$ is a nonzero integer if and only if
- $n$ divides $1$
- either $n-1$ divides $1$ or $n+1$ divides $1$, and
- there is a power $x$ of $t$, so that $\dfrac{x-1}{t-1}\equiv n \pmod {t-1}$


Is the proof as follows?

From an other lemma we have that $(t^n-1)/(t-1)\equiv n \pmod {t-1} \tag {1}$.
 
Hi @Ted.
 
5:06 PM
Oh, actually the proof of 3-manifolds appearing as branched covers of S^3 can be done in a staightforward way. Take your 3-fold $M$, Heegaard decompose to get two handlebodies. Now these handlebodies appear as branched cover of $D^3$ branched over $g+1$ arcs. Now one has to check if the way handlebodies above are glued will ensure you get a branched cover of $S^3$ below.
 
@DanielFischer @robjohn @anon @SamuelYusim Could you take a look at my proof above and tell me if it is correct or if I could improve something?
 
I sketched this for you. I still don't know how to show that those handlebodies branch over $D^3$.
 
Using the same method I mentioned above.
 
OK, got it. Thanks.
 
i.e., take an axis passing through the handlebody. It hits the surface in $2g$ points. Now flip along that axis. The $g$ arcs inside the handlebody are left fixed under this action. Delete these and quotient. You have a cover over $D^3$.
@MikeMiller if you're not interested, I won't bother you.
sorry.
 
5:09 PM
I edited the comment precisely because it gave the wrong tone.
Sorry about the first version.
I see now. Thanks.
But it looks like you're describing a degree 2 branched cover. Not every 3-manifold can be obtained by a degree 2 branched cover, so you're going to run into trouble when trying to glue this along the boundary
 
@lopata $4\,\vphantom{F}_2F_1\left(\frac12,1;\frac32,-1\right)$
 
yeah, you're right. let me ponder on how to finish this.
 
5:29 PM
@DanielFischer Why is it that in a Banach algebra it doesn't follow simply that $E^{2}-E = 0 \implies E(E-1) = 0 \implies E = 0~~\text{or}~~E=1$. Since a Banach algebra is an algebra over a field, the multiplication is distributive, I don't see why this doesn't follow?
 
5:55 PM
@Moses: You're assuming there are no zero divisors, but that doesn't need to be the case.
 
@MikeMiller I'm not following. Could you give some detail please?
 
@RandomVariable Depends. If - as it is most likely the case here - it's just somebody trying to be nice to you, there's no need to report it, just be prepared to see it reverted tomorrow. If it happens more often, and you are annoyed by the repeated reversals, notify us, so we can have a friendly word with the serial upvoter.
(Serial upvoting is, with the obvious exception of sockpuppeting or voting rings, nothing really bad, so it's unlikely to lead to a suspension. But of course the reversals tend to be annoying for the target, so a friendly reminder that they should vote un-serially would be sent.)
 
@Moses: I mean, let's cheat. Consider the $k$-algebra defined by $k[x]/(x^2-x)$. This certainly has a nontrivial element such that $x^2-x = 0$.
 
@Moses Consider $\mathbb{C}^2$ with componentwise multiplication. Then $(1,0)^2 = (1,0)$, but $(1,0)$ is neither the unit - which is $(1,1)$ - nor $0$.
 
Alternatively if $k$ is any field the algebra $k^2$ (with unit $(1,1)$) has four idempotents.
Boo.
 
6:02 PM
@DanielFischer So what additional assumption do you need to have on the space in order to have that $E^{2} - E = E(E-1) = 0$ implies $E = 1$ or $E = 0$?
 
@Moses You need to have no zero divisors. The algebra being one-dimensional would give you that. I'm not aware of other general sufficient conditions. But it's of course not necessary, there are infinite-dimensional Banach algebras without zero divisors.
 
@DanielFischer: A commutative algebra has nontrivial idempotents iff it splits as a product $A \times B$. But of course commutative Banach algebras are not the norm.
 
@MikeMiller Yes. But can you characterise splitting in a nontrivial and useful way? I'm not aware of a good criterion.
 
Define useful. I find the above to be a very satisfying equivalent condition.
I guess "something you can check by hand", but then I'm also not sure what by hand means. (I guess for an algebra you've 'given me' I probably will have an easier time finding idempotents than splittings, but I think the notion of the splitting above tells me quite a lot about the structure of our algebra.)
 
@DanielFischer You said you need no zero divisors. What if you have zero as the only zero divisor. Such as in $\mathbb{R}$?
 
6:16 PM
@MikeMiller Hmm, I guess in practice it would usually be pretty easy to see whether you have idempotents or not (I see you also guess that ;). I kind of looked for a condition that is not directly equivalent, perhaps sufficient but not necessary, and that would be easy to check.
 
@Moses: an element $e$ such that $e^2 = e$ is called an idempotent. Having no zero divisors is sufficient to knowing that there are no nontrivial idempotents (if $e$ is an idempotent, then $e(1-e) = 0$, as you noticed, so if there are no zero divisors $e=0$ or $e=1$) but it is not necessary.
 
@Moses In that case, from $E(E-1) = 0$ you can deduce that $E = 0$ or $E - 1 = 0$.
 
In the algebra $k[\varepsilon]/(\varepsilon^2)$, there are no idempotents - $(a+b\varepsilon)^2 = a^2+2ab\varepsilon$, which is $a+b\varepsilon$ iff $b=0$ and $a$ is zero or one - but of course it has a zero divisor, $\varepsilon$ itself.
 
@MikeMiller Yeah but I was just stating that that would also hold with the statement that $0$ is the only zero divisor. You would still get that $E = 0$ or $E-1=0$.
 
 
1 hour later…
7:47 PM
@MikeMiller Could I ask for a hint regarding something?
 
8:01 PM
i always used to struggle for understanding fuzzy logic, i think there is someone who can demystify it for me :D
right @NeuroFuzzy ?
 
@Agawa001 :D sorry, my expertise only extends to rice makers.
 
oh, alas...
 
@Moses: You can try, but I don't know terribly much about operator algebras.
 
@MikeMiller I've only starting to learn it recently so you probably at least know terribly much more than me...I had a moment of weakness. Going to try to figure it out without a hint...for a while longer.
 
8:22 PM
Whoa, Tim Burton's Frankenweenie is actually pretty good.
 
Mhm.
Oh, there's a 2012 one? I saw his 1984 short film.
 
I didn't know there was a 1984 short film.
 
I assume this is a remake of that.
 
@lopata made something that would shake the brain of many mathematicians
 
i'm amused that both of you initially thought you were talking about the same thing :P
 
8:25 PM
@Semiclassical: I mean, there are two things with the same name, so I don't think that's too surprising.
 
oh, of course not. it's just kind've interesting from a philosophical perspective
though i guess it'd be more so if that conversation went on longer
i.e. each participant being able to engage in conversation without actually having the common ground they think they have
 
I especially like the quality of the (2012) animation. I've only seen one animation which is something close to this, and that's Coraline.
 
@Semiclassical Ever listened to a conversation between a physicist and a mathematician?
 
hah, yep
 
lol
 
8:30 PM
reminds me of an old 'test' for distinguishing between a physicist and a mathematician
$$f(x,y)=x^2+y^2\implies f(r,\theta)=?$$
 
you mean $f(r,\theta)$, or at least that's what you showed me before
the version with the comma was pretty funny and actually works
 
yeah, i was fumbling with mathjax
 
you really tested it on somebody?
 
@lopata the numerator of your picture is true thu , it has imaginary peer
 
sure, why not
 
8:32 PM
eh, i forget if i have. but it's not meant as a literal test, more of a 'how would they react'
 
Howdy @Semiclassic. Goodnight, @Balarka, MikeM
 
though for some it probably would literally work :P
afternoon @ted
 
it literally does
 
now you're doing that goodnight thing with me too, @Ted?
hi, btw
@MikeMiller haha, wow
 
No, Balarka, it's past your bedtime!
 
8:33 PM
@Ted: Ciprian approves of the thing I sent you (plus some typos I fixed), but I'd still like your opinion.
 
my own reaction when i saw it was "well, it's ambiguous, isn't it?" which i chalk up to being a mixture of both math and physics
 
I think it's good, Mike, but if you can make it slightly less background and more about your envisioned project, that would be stronger.
 
OK. Thanks.
 
@TedShifrin When isn't it past my bedtime?
 
Research proposals are a tricky thing.
 
8:36 PM
The idea is fairly straightforward, so the change would be to spend more time on the actual technicalities I need to overcome.
I think the word "I" is in there twice, so it's fair that I didn't talk much about the actual project.
 
i think you make it stronger by showing you have actual specific ideas in mind.
 
@Semiclassical a funny question (not sure you saw it) :-) $$\lim_{n\to \infty} \left(\frac{n}{\log(n)}\right)^2 \int _{\large 1/n^{1/n}}^1\int _{\large 1/n^{1/n}}^1\frac{1}{(x+y) (1+x y)} \ dx \ dy$$
 
what kind've idea are you after, out of curiousity? @MikeMiller
@Chris'ssistheartist ...that's kind've terrifying
 
@Semiclassical It's extremely easy, it just seems tough but it's not. :D
 
ah. so it's only scary-looking
 
8:41 PM
Exactly, it's deceiving.
 
i suppose what i'd do first is substitute $\epsilon=n^{-1/n}$ so that the argument becomes $$\frac{1}{(\log \epsilon)^2}\int_\epsilon^1 \int_\epsilon^1 \frac{dx \,dy}{(x+y)(1+x y)}$$
 
@Semiclassical That looks good. :-)
 
just to make things look nicer and to consider $\epsilon\to 0$
woops, yes. $\log \epsilon\to 0$.
 
Yeap. I had in mind another problem I created these days.
 
$\chi_2(\cdot)=?$
 
8:47 PM
@Semiclassical Legendre chi function.
 
how does that affect on the integral
 
ah. haven't worked with it myself
 
@Semiclassical: Use an enhanced version of the Seiberg-Witten invariants to study the topology of the group $\text{Diff}(M)$ for certain 4-manifolds.
 
@TedShifrin Fair enough.
 
8:47 PM
ahh, sounds interesting
what kind of 4-manifolds?
 
It's not really what kind, it's that for specific families you have enough diffeomorphisms to play with by construction.
 
@robjohn the question I deleted above is very interesting for some reasons.
Let me check a bit the books on Amazon and see if new books on integrals, series and limits were published in the meantime ... :-)
(there is some stuff on infinite series)
 
out of curiosity, and having nothing better to do right now, how does so much physics enter into the play in low dimensional topology?
 
With gusto.
 
:P i wanted something a bit more philosophical than that.
 
8:57 PM
i'm a little curious as well, though i think i'm more intrigued by what physics structures tend to enter in
which, given the reference to seiberg-witten, probably means quantum field theory
 
ahh, nice
i'm not someone who really understands SUSY+QFT from the physics side, alas, so i'm useless in either side of this
 
what's an instanton?
 
if you're asking in quantum mechanics, i can give an answer. not in terms of topology, though
 
if you can put it in a few words, i'm prepared to listen, though i can't guarantee you i'll be keeping up with you
 
9:01 PM
fair enough. hmm
 
You don't need to know what that is to read that article.
 
Donaldson shocked the topology world
 
probably true. i'm mildly curious since i have heard the name sprout up in places
 
here's a very simplified statement in relation to the physics
and only by analogy
 
I bet there are very nice books about integrals, series and limits that some authors prefer to publish in their countries only. Let me know if you are aware of such books in your countries (I wanna buy some).
 
9:05 PM
suppose i have a system which has two preferred configurations. say, it prefers to point straight up, or straight down. one can describe other states, but they won't be stable
now, in classical mechancis, if it's in one of those states, and it doesn't have a lot of energy, then it'll stay in that state. (if it had more energy, it might be constantly changing between them, like a compass needle constantly spinning. but i don't want that case.)
so it either points up, or it points down.
 
ok.
 
in quantum mechanics, though, there are reasons why that doesn't hold. one way to say it is that there are (approximate) solutions of the following kind: Start out pointing up, stay that way for a long time, then suddenly flip to the other and stay that way.
by doing it quickly, you minimize the time you stay in that unstable region.
 
mhm
 
usually it's more subtle than that, with there being a cost to changing quickly. in that case you usually end up with something like $x(t)=\tanh(t/t_0)$
with $\pm1$ being flipped up/down
and with $t_0$ being some very short time scale
now, in QM, you're usually just dealing with a finite number of degrees of freedom. so you could have something more complicated than just up/down
in QFT, though, you moreover talk not about particular degrees of freedom (say, where a particular particle is) but in terms of quantum fields
 
you're losing me, but go on.
 
9:13 PM
so you instead have objects like $\phi(x)$, the strength of some scalar field $\phi$ at each point in space
 
ok, yes.
 
so now instead of an individual object flipping up/down, you have to envision that entire field having that kind've character
in which case you might end up with something like: "On the LHS of the system, all the electrons point down; on the RHS, they all point up. In between, the local orientation changes as $\tanh(x/L)$."
 
sure
 
that's more an 'instantaneous' change in space than in time, though. i know there are also field configurations that change in time, but i don't do QFT enough to really appreciate it
and i suspect that's problematic when it comes to Seiberg-Witten stuff, since that really is about (supersymmetric) field theories
 
ok, totally lost now. anyway, thanks.
i've to go and get some sleep.
g'night.
 
9:20 PM
It looks like my math thesis is officially going to be over a topic in nonstandard analysis :o (undergraduate senior thesis)
 
@BalarkaSen night
 
i don't know enough math to write a senior thesis :(
 
[looking at that arxiv reference] "oh god prepotentials"
 
@Semiclassical: I think that paper should tell you the correspondence between the physics and the math, something about S-duality I think. I've not actually read it.
 
yeah, section 11 looks vaguely familiar to stuff i read two years ago
the thing i'd be curious about is if there's relations with resurgence theory, which also deal with field theory and instantons (and which I know Witten is aware of)
 
9:30 PM
¯_(ツ)_/¯
 
nods
i don't suppose the name Nekrasov is at all familiar to you?
 
nope
 
mmkay
 
keep in mind I know literally none of the relevant physics
 
nekrasov is a big name to some of the folks i know on the field theory side. seems to be related to the notion of 'instanton-counting' which i'm utterly ignorant of
 
9:42 PM
Hi i'm not sure if this q'n is worth an actual question so I'll ask here first: If the chance of winning is 44% and if i win there's a 4% chance of winning $2 otherwise $1. What's the overall chance of winning $1?
 
@MikeMiller the first page of these notes seems to give a summary, understandable or not
with seiberg-witten invariants being discussed in the first indented paragraph
 
I'll have to look another time. I'm on my phone, can you save that link for me for later?
 
sure.
for me the barrier is that the SW stuff, as physics, is to do with 4D field theories, and i just don't work with that stuff
plus, y'know, supersymmetry
 
dunno either of those :) give me until I finish my PhD
 
9:58 PM
heh. well, seiberg-witten stuff is about N=2 supersymmetric field theories (whatever the hell that is)
anyhow, i should head out. good luck with your proposal
 

« first day (1856 days earlier)      last day (3164 days later) »