« first day (1408 days earlier)      last day (3614 days later) » 

12:00 AM
@KevinDriscoll Thanks. What is the intuition behind Laplace Transforms? What do they do to a function? I don't understand why the condition of decay (ie, function must be of exponential order) is needed.
 
@masfenix Well if it grows faster than exponentially then the transform doesn't exist for any $s$
 
@KevinDriscoll YO YO
 
The intuition though is that the Laplace domain is one in which differentiation by $t$ is mapped ot multiplication by $s$ @masfenix
What it does to functions though..... I can't say for sure. It's more clear with Fourier and Mellin. I don't work much with Laplace @masfenix
 
Oh, I think I get it now. The decay condition is needed for the integral to converge. If the integral dosn't converge then obviously we don't have a Laplace transform.
 
Indeed.
@PedroTamaroff Yo! Good to see you.
 
12:06 AM
@KevinDriscoll What's up?
 
@Pedro Working on my PhD thesis proposal. And crying on the inside at how few jobs are available in my field.
 
@KevinDriscoll Ah, cannot be less than those for a mathematician, for sure!
 
@Pedro You'd be surprised. The problem is physics is more segmented. Whether you do theory or experiment you're automatically disqualified from the other's job postings.
I think there are also fewer math PhDs. But I know its also pretty bad over there. From my cursory estimates and personal experience I'd say its about even.
 
@KevinDriscoll Hehe, I'll tell you when I graduate.
 
@pedro I'm hoping to not wait until I'm ready to graduate to really be looking for things
 
12:22 AM
@KevinDriscoll But you're already a PhD student right?
@AlexanderGruber HAI.
 
@PedroTamaroff Bon soir.
 
@Pedro Yeah, so a good bit farther along. Seems like a lot of people don't start looking for jobs until they graduate tho.
 
@AlexanderGruber What have you been up to?
 
@Kevin oh you're a physics PhD student?
@PedroTamaroff torturing my business calc students :)
 
@AlexanderGruber Yup. Cold atom theory.
 
12:24 AM
@KevinDriscoll sounds fun. i like physics, did my undergrad in it
i'm not quite awake yet.
 
@AlexanderGruber Did you also do math?
 
@KevinDriscoll yeah, doubled up
 
AH okay. I wish I had done math sometimes.
 
@KevinDriscoll did you during undergrad?
@PedroTamaroff yourself?
 
Nope. I was a physics/philosophy double @alexander
 
12:27 AM
@KevinDriscoll Ahhh interesting.
So you can ponder the ethics of relativity
 
Hahahaha. Well it all depends on you frame of reference really.......
2
 
@AlexanderGruber Algebrain' pretty hard.
I failed my combinatorics test as I expected. =P
 
@pedro you failed a test on combinatorics? Howd that happen?
 
@KevinDriscoll A series of bad choices and mistakes during the exam.
 
@pedro What topics did it cover?
 
12:38 AM
Elementary Counting, Generating Functions, Graph Theory, Incidence Algebras of posets.
 
@pedro I probably would fail that test too. Generating functions are aparticular weak point for me. Though we use a generalized version in Quantum Field Theory
 
 
5 hours later…
5:25 AM
@Hippalectryon $\mathbb Q$ Lacks Bolzano-Weierstrass property
 
 
1 hour later…
6:33 AM
@mick I can't talk about magic floopskintunks and say that they make sense but I can't explain, right?
First show me how you define exp(X) in R[X] modulo X^3 - 1. You are not even sure whether or not it makes sense, i.e., is a polynomial.
Hey @Alexander.
 
hi
real empty in here tonight
 
empty?
 
 
Heh.
I couldn't find any stuff in Isaacs, by the way @Alexander
 
@Balarka Sen, I've been on a quintic trip for the past few days thanks to you. Teehee
 
6:42 AM
@DavidKirby Oh? Tell me about it.
@Alexander I totally not approve the weird man mustache.
 
@BalarkaSen hahaha
i grew it out specifically for my ID picture
i don't normally have facial hair
 
@Balarka Sen, well, I've been working a lot with Harmonic Series lately and have come up with notebooks full of interesting series. After talking about quintic equations with you I started trying to solve various series with a quintic equation on the right hand side, as well as trying to rewrite different types of series in quintic form and the results of have been really interesting.. Roots galore, which I guess is to be expected...
 
@Alexander To be a mathematician, you need to look like this
bad pun
@DavidKirby Examples?
 
@Balarka Sen, sure, gimme a sec and I'll type in an example
Ok, here's one -- $$H_{-x-\frac{3}{4}}-H_{-x-\frac{1}{4}}=x^5-x-Z$$
Solving for Z gives a really interesting result :
$$Z\to \frac{1024 H_{-x-\frac{1}{4}}-1024 H_{-x-\frac{3}{4}}-1024 \left(-x-\frac{1}{4}\right)^5-1280 \left(-x-\frac{1}{4}\right)^4-640 \left(-x-\frac{1}{4}\right)^3-160 \left(-x-\frac{1}{4}\right)^2+1004 \left(-x-\frac{1}{4}\right)+255}{1024}$$
 
Whoa.
That's odd.
 
6:56 AM
yeah now check this :
$$H_{-x-\frac{3}{4}}-H_{-x-\frac{1}{4}}=\frac{-1024 H_{-x-\frac{1}{4}}+1024 H_{-x-\frac{3}{4}}+1024 \left(-x-\frac{1}{4}\right)^5+1280 \left(-x-\frac{1}{4}\right)^4+640 \left(-x-\frac{1}{4}\right)^3+160 \left(-x-\frac{1}{4}\right)^2-1004 \left(-x-\frac{1}{4}\right)-255}{1024}$$
solving for x :
$$\{\{x\to -1\},\{x\to 0\},\{x\to -i\},\{x\to i\},\{x\to 1\}\}$$
 
Ah, so you have a solvable quintic up there.
Let me understand what you are trying to say,
$x^5 - x - \tau = \mathcal{H}_{-x-3/4} - \mathcal{H}_{-x-1/4}$
When $\tau$ is the humongous thing above, right?
 
Correct, ... my thinking was that this is the expected behavior though, right?
 
So shouldn't the harmonic number terms cancel out, leaving you some algebraic identity?
 
Yup, it reduces to : $$-H_{-x-\frac{1}{4}}+H_{-x-\frac{3}{4}}-x^5+x$$
 
How do you find it interesting?
 
7:06 AM
Well, it just hit me by surprise, and it's a relationship that I only understand in a one-sided sense.. I think I need to study up on Galois groups (and cyclic permutations especially) .. Not sure where to begin
 
How do you even get that identity? Seems weird to me.
I never imagined a connection of quintics with harmonic numbers.
Could harmonic numbers then be inverted to solve a quintic?
 
Heh, neither did I.. It was the weirdest thing, tried it on a whim after talking to you the other night.
 
Like, asking to solve $\mathcal{H}_x = \tau$ for $x$.
 
@Balarka Sen, I was wondering that, but hadn't figured out how to go about it yet
 
How is $\tau$ related to $x$ in the above quintic identity?
$$x^5 - x - \tau = \mathcal{H}_{-x-3/4} - \mathcal{H}_{-x-1/4}$$
@David
 
7:16 AM
like this? $$\tau =H_{-x-\frac{1}{4}}-H_{-x-\frac{3}{4}}+x^5-x$$
uhm, err
and solving for $\tau$ gives you that big equation
and the big equation reduces back down to $$\tau =H_{-x-\frac{1}{4}}-H_{-x-\frac{3}{4}}+x^5-x$$
it's got my brain all blended for sure heh
 
Ah. But that's not weird at all, is it?
We just have another algebraic form for $x^5 - x$
And the solutions you get for $x$ are really the solutions of $x^5 - x = 0$
 
Yup, -- so that definitely makes sense.. Ok, so I'm still a little baffled by the solution after substitution :
$$H_{-x-\frac{1}{4}}-H_{-x-\frac{3}{4}}=H_{-x-\frac{1}{4}}-H_{-x-\frac{3}{4}}+x^5-x$$
$$\{\{x\to -1\},\{x\to 0\},\{x\to -i\},\{x\to i\},\{x\to 1\}\}$$
so instead of the 5 roots of the quintic, you get that
 
Yes, the roots of $x(x^4 - 1) = 0$, the simplest kind of quintic you can think of.
 
AH
Hah, ok it all becomes clear to me now
... I have fooled myself with slight-of-hand
 
7:34 AM
I am trying to prove that a quintic cannot be solved in terms of polygamma and polylogarithms, based on Murty-Saradha-Gun conjecture about their algebraic independence.
Essentially tracing steps of T. Chow, who proved it for exponential and logarithms.
(assuming Schanuel's conjecture)
 
If so, it could be a boon to transcendental number theory
 
Probably. I've never heard of any result regarding the fact that polylogarithmically-closed of polygamma-closed fields only contain algebraic with solvable galois group.
 
7:53 AM
David, I have a probability question I think you will enjoy
I am enjoying it, at least
 
8:10 AM
Hrm, unsure..
 
@Studentmath, .... cuuuurrrious
 
You have a box with 6 coins, numbered 1 to 6. Each has a probability of 1/i (where i is its number) to land Head up. You have a fair dice, with 6 sides.
You throw the dice once, pick up the coin according to the number that it landed on, and throw it until you land Head up.
Let X be the RV representing the result of the dice, and Y be the RV representing the number of times you throw the coin until (including) it lands Head up.

E[Y], Var(Y), and $P(Y\ge 5$| X is even)
It's really nice, you get to toy with few known probabilities but in a nice controlled and adjusted enviroment
 
8:29 AM
Greetings
@robjohn Is there a known form of $\zeta''(2)$ expressed in terms of known constants?
@robjohn I wonder what is the best place to publish such a discovery (just in case).
 
@Chris'ssis I think so... let me look
 
8:47 AM
@Chris'ssis Heh... I entered Sum[(Log[k]/k)^2, {k, 1, Infinity}] into Mathematica and got a useless result :-)
 
@Studentmath, interesting.. i gotta jet right now, but i'll be back later on this morning and will tinker with that a bit more
 
@robjohn I have to go now. Pls let me know if you find something about it.
 
9:35 AM
@Chris'ssis Unlikely.
 
@BalarkaSen A friend of mine said he discovered such a thing ... (I need to see to believe)
 
$\zeta'(2)$ has a closed form though. Interesting question.
@Chris'ssis Try arxiv for a preprint, as an initial publication.
 
@BalarkaSen If someone like me discovered such things, most probably would be accussed by intellectual property theft or something like that. I can guess many would claim it's their discovery (since I have no math background).
 
@Chris'ssis Your works are original, no? So why would anyone accuse you of thieving?
 
@BalarkaSen Hypothetically speaking ...
 
9:41 AM
You're being silly.
Nowadays, it doesn't matter if you don't have a math background.
People publish there work and none will claim it's theirs.
 
@BalarkaSen Maybe ...
 
If your works are original and of general interest, publish them!
It's far better to publish research works than publishing a contest book.
@robjohn
Given any formal power series $f_1,...,f_n$ in $t\Bbb C[[t]]$ which are linearly independent over $\Bbb Q$, then the field extension $\Bbb C(t,f_1,...,f_n,\exp(f_1),...,\exp(f_n))$ has transcendence degree at least $n$ over $\Bbb C(t)$. [Ax-Schanuel theorem]
What the heck is $t\Bbb C[[t]]$?
 
@BalarkaSen I assume complex power series without constant term
 
@robjohn $\Bbb C[[t]]$ is the ring of formal power series, but I don't understand what's the $t$ sitting up in the front.
 
@BalarkaSen it means multiply by $t$ to get rid of the constant term
 
9:51 AM
Weird notation for that.
@robjohn OK, thanks. That makes sense.
But wait a minute. The aren't $f_i$s linearly dependent over $\Bbb C(t)$? All having a factor of $t$?
 
Heyall
 
Ahoy.
 
How can I use that $\sqrt{1-t} \leq 1-t/2$ holds for $t \in (0,1)$ to prove that
$$\frac{1\cdot3\cdot5\cdots(2n-1)}{2\cdot4\cdot6\cdots2n}\geq\frac1{2\sqrt n}$$?
 
Yuck.
Back to Ax-Schanuel : Doesn't the linear independence of $f_i$s over $\Bbb C(z)$ mean we are excluding the $n$-tuple candidate $\{f_i\}$ outright?
Plus, excluding all $n$-tuples with multiple $f_i$s in it?
 
1
Q: Using $\sqrt{1-t}\leq 1-\frac t2$ to show that $\frac{1\cdot3\cdot5\cdots(2n-1)}{2\cdot4\cdot6\cdots2n}\geq\frac1{2\sqrt n}$

AleksanderI have a problem that tells me to use that $\sqrt{1-t}\leq 1-\frac t2$ for $t\in(0,1)$ to show by induction that $\frac{1\cdot3\cdot5\cdots(2n-1)}{2\cdot4\cdot6\cdots2n}\geq\frac1{2\sqrt n}$ So far I have shown that $\frac{1\cdot3\cdot5\cdots(2n-1)}{1\cdot2\cdot3\cdots n}\leq 2^n$ from a previou...

 
10:06 AM
Cheers
 
Should I jettison the dead weight then (i.e., @N3buchadnezzar)?
 
@N3buchadnezzar We are in a pirate ship. Didn't I say "Ahoy"?
 
@N3buchadnezzar I think you wanna see my proof ...
 
@BalarkaSen yarr
 
10:09 AM
That sum looks familiar.
Note that the identity below $(3)$ is weirdly resembles the Chudnovsky brother's formula, @Chris'ssis @N3buchadnezzar
These are called Ramanujan-Sato series. Who knew?
So rigorous, much results, very identities, wow.
 
There has to be a simpler way to do this.. sigh
Any probability lads around? Nothing too complicated, just trying to figure out a shorter way to do something
 
10:29 AM
@Chris'ssis Wiki says Ramanujan knew something to compute these series we don't. Interesting, isn't it?
 
@BalarkaSen Indeed. (I hope to be like Ramanujan one day)
 
@Chris'ssis There are claims that he had alien help :-)
4
 
I reckon Ramanujan must've thought about modular forms and the surrounding algebraic theories long before we did.
 
@robjohn lol :-)
 
As I always say : Ramanujan is the father of modular forms. He knew much than we think he did.
Perhaps he created them exactly with the same rigor we introduce these days.
 
10:31 AM
@DanielFischer That's funny, the exact same guy asked the same question on a different site. (Norwegian). Seems the problem with your explenation (which I found exellent) is that he seems to miss the crucial fact that $\prod_{k=1}^n k =\prod_{k=2}^n k = n!$
 
Hmm. How?
 
@Chris'ssis Well, it's never late to be like a classical hero in mathematics.
 
True
 
@DanielFischer Well it seems he does not understand why the product equals $1/\sqrt{n}$
 
I sure do hope to be like Erdos someday. But alas, I don't think I can ever be like him.
 
10:34 AM
I hope to be like Galois one day
 
@N3buchadnezzar Yes. I wonder how that is possible.
 
@N3buchadnezzar Whoa.
He was another guy like Ramanujan.
 
Swimming in p***y, and death by manly duel + mathematician.
 
@N3buchadnezzar Pah, better be like Ian Curtis. Or the Lizard King.
 
@N3buchadnezzar Ah.
@DanielFischer Lizard? Curtis Connors?
 
10:37 AM
@BalarkaSen James Douglas Morrison
 
Don't know who that is.
K, I gotta go.
 
later pal
 
@DanielFischer There was an idea of making a nirvana guitar hero game, where the controller would be the shape of a shot gun..
 
..not exactly sympathetic imo
 
10:45 AM
@N3buchadnezzar That could be considered a little in bad taste.
 
Yeah, although I would not say no to more nirvana
 
sometimes suicide is the only way to relieve the pain.
 
@robjohn the solution to the integral involving the dilogarithm
 
@Chris'ssis why are you writing $i^2$ instead of $-1$?
 
@robjohn it looks nice that way ( the negative result annoys me)
 
10:52 AM
@Chris'ssis Why? Haven't you accepted negative numbers yet :-)
 
@skullpatrol :-))))))))
 
@skullpatrol you've heard of the double negative, this is the double imaginary.
 
I do not believe in negative numbers
 
@N3b what about non-integers?
 
10:54 AM
@DanielFischer I feel so old.
 
@robjohn Why that?
 
@DanielFischer people here don't know about The Doors...
 
@N3buchadnezzar Do you believe in zero as a number?
 
10:57 AM
@robjohn That's not age, that's lack of education. Like if they didn't know about Marlowe (both).
 
< 3 Riders on the Storm
 
@DanielFischer It has to do with when people grew up and what music was playing at the time.
 
I grew up with hip hop nods
 
@robjohn Good music is timeless.
 
There has to be an easier way to do Var(Y), I can't believe I actually have to do all these $Cov(Y_i, Y_j)$. I'd like to think they are all zero, but I can't find a good explaination. So I guess they aren't. Which means, I am stuck.
 
11:00 AM
@DanielFischer Cream?
 
They are exclusive, so they can't be independent..
 
@N3buchadnezzar Not so much my thing. Good, but not with the necessary je ne sais quoi.
 
Creedence?
 
When people say "grew up with;" that means, to me, the popular music when they were 13 to 19 years old.
 
11:02 AM
@N3buchadnezzar Definitely.
 
@skullpatrol That's about right, but for me, I think it was until 26 (I listened to and still really enjoy music from when I was in grad school and a couple of years after)
 
Indeed, music is a very personal taste...
...like the music of the spheres :D
 
Hello there !
 
11:07 AM
Hi!
 
@robjohn in my proof there is a text that I forgot to remove ... "By letting $x\mapsto -x$ in $X$, we get"
 
http://math.stackexchange.com/questions/397097/uniform-convergence-of-sum-n-1-infty-frac-sinn-x-sinn2-xnx2/398085#398085 why do we have $\begin{align}
\sum_{n = 1}^N{\sin{(nx)}\sin{(n^2 x)}\over n + x^2} & = \sum_{n = 1}^N {1\over n+x^2}(S_n - S_{n-1}) \\
\end{align}$ ?
($S_n = S_n(x) = \sum\limits_{k = 0}^n \sin{(kx)}\sin{(k^2 x)}$)
 
It may have something to do with $\sin(a)\sin(b)=\frac12(\cos(a-b)-\cos(a+b))$
 
@Hippalectryon Isn't it related to the summation by parts?
 
@Chris'ssis it is
 
11:14 AM
@Hippalectryon OK :-)
 
@Chris'ssis but it comes after that part
 
Music is noise.
 
@BalarkaSen how long ago did you sell your soul? ;-)
8
 
a couple of years ago, i believe.
 
Don't worry you can get a cheap new one on ebay :3
 
11:17 AM
wait, what?
@Hippalectryon this is star-worthy
 
:)
'Mathematician looking for a soul - cheap - good quality - science resistant - contact me anytime'
 
@Hippalectryon no denomination specified?
 
Who starred?
You fool.
 
I did.
 
@ParthKohli I know you did.
 
11:20 AM
@BalarkaSen only robjohn's.
 
@robjohn But a zero denominator is not allowed ;-)
 
Wait, 3 stars?
 
@robjohn I'm not used to doing this - i already stole my 30 classmate's souls :D
 
@Hippalectryon And now you're selling them cheap?
 
@BalarkaSen gotta keep the product moving
 
11:22 AM
Just don't make a horcrux outta those, that's all.
 
@BalarkaSen Exactly, i need money for an ice cream :D
Once i have my ice cream, i multiply it into 2 ice creams thanks to the axiom of choice
And therefore i have infinitely many ice creams
 
No, it's Banach Tarski.
 
Isn't it based on the axiom of choice ?
 
yeah, but it's nonobvious.
 
Hey i'm not telling how i multiply ice creams to everyone :P
 
11:23 AM
Haha
 
If I take your ice cream away; is that negative ice cream?
 
@skullpatrol :O Don't ! nuuu my ice cream - hey wait, i'll just need to buy something to take the absolute value of ice creams :D (Unless it creates negative entropy duh)
 
You should bye a comathematcian, @Hippa, if you're interested in turning cotheorems into ffee.
 
@BalarkaSen HTCPCP is enough for me :)
 
11:27 AM
The Hyper Text Coffee Pot Control Protocol (HTCPCP) is a facetious communications protocol for controlling, monitoring, and diagnosing coffee pots. It is specified in RFC 2324, published on 1 April 1998 as an April Fools' Day RFC, as part of an April Fools prank. An extension is published as RFC 7168 on 1 April 2014 to support brewing teas, which is also an April Fools' Day RFC. Protocol RFC 2324 was written by Larry Masinter, who describes it as a satire, saying "This has a serious purpose – it identifies many of the ways in which HTTP has been extended inappropriately." The wordi...
Is the Banach Tarski's 'paradox' related to Cantor's bijection between lines and areas/volumes ?
 
Q : Why did Bourbaki stopped writing books?
A : Lang was only one person.
 
what is $||A||$ for $A\in\mathbb{R}^n$ ?
 
@Hippalectryon What is $||\cdots||$?
 
11:42 AM
Is it $\sqrt{\sum\limits_{k\in A}k^2}$ ?
@BalarkaSen The norm i think ... not sure
 
Maybe. Euclidean norm, you mean?
 
In a correction of an exercise I have (with some conditions before)
A is a square matrice and $X\in\mathbb{R}^n$
 
Is X also a matrix?
ah.
Then it's Euclidean norm.
 
That is .. ?
 
square root of sum of squares blah
 
@usukidoll lol ?
 
what kind of question is that?
does it even have a question?
 
JMK
@usukidoll looks like the OP must learn some typesetting
 
@robjohn have you tried this one? I plan to compute it now. $$\sum_{n=1}^{\infty} (-1)^{n+1 }\frac{1}{n} \left( 1+ \frac{1}{2^2}+ \cdots+\frac{ 1}{ n^2}\right)= \zeta(3)-\frac{\log(2)}{2}\zeta(2)$$
 
11:54 AM
I actually went apes*** on someone because the op was taking an exam and posting his test paper on here
it was in another language
anyway, anyone on here know partial differential equations?
 
JMK
@usukidoll That's an irony, I was about to ask the same thing
@usukidoll What is the problem?
 
well I was just practicing but I think my outline book is making errors yo
 

« first day (1408 days earlier)      last day (3614 days later) »