« first day (1506 days earlier)      last day (3506 days later) » 

12:20 AM
Off topic: Does anyone have references on analysis (at the level of royden) and measure theory? I'm looking for books or sources that have lots of exercises.
 
Stein/Stakarchi, Folland ?
 
1:09 AM
@Hippa: smacking $\ne$ bashing!
 
1:22 AM
If $f(x) \sim g(x)$, then is $\lim_{x\to 0}f(x) - g(x) = 0$?
 
1:50 AM
@TheSubstitute I'm using the book of Dudley and is very beautiful. In particular I fall in love with the proof of Tychonoff´s using ultrafilters. Also I use when is possible the books of Tao.
 
2:23 AM
If someone have time, maybe can check the following math.stackexchange.com/questions/935693/…
 
2:57 AM
@VibhavPant No. For example, $x \sim \sqrt{2x^2}$ but the difference gets arbitrarily large.
 
 
3 hours later…
6:26 AM
Man, this edit from Felix is not only useless it makes my latex look worse. Sigh
4
Q: Closed form for $\int_0^x \{1/t \}\,\mathrm{d}t$, $x \in \mathbb{R}_+$ and related.

N3buchadnezzarAfter some tests I think that Conjecture 1 Let $x \in \mathbb{R}_+$ then $$ \int_{0}^{x}\left\{\,1 \over t\,\right\}\,{\rm d}t = 1 - \gamma + H_{\left\{1/x\right\}} - x\left\lfloor\, 1/x\,\right\rfloor + \log\left(\,x\,\right) $$ Where $\{x\} = x - \left\lfloor\, x\,\right\rfloor$ is the...

 
Why does mathjax freeze my chrome browser?
 
6:54 AM
Arthur Fischer mentioned in this post that: In the last 30 days well more than 1000 comments have been removed by the moderators. This just reminded me how much work moderators do on this site. (In particular Arthur who is at the moment most active mod on meta.MSE.) So I wanted to publicly thank them for their work.
3
I could have posted this as a comment on his post. But I did not want to bother him with an unnecessary ping. And, more importantly, I thought that here it would be more visible than in a comment there, which would very probably go unnoticed. Not to mention that the comment is not that relevant to the post and, knowing Arthur, my guess is that he would use his mod powers and after a few days he would delete a comment praising him as being obsolete.
(If you can think of some more suitable place where something like this - appreciation of some MSE user and their work/achievements - could be posted, feel free to let me know.)
@IceBoy For me the questions on main are displayed correctly in Chrome (37.0.2062.120 m). Admittedly, this information is probably not much help to you :-(
 
@MartinSleziak thanks for the info, any help is appreciated :-)
We should, in my opinion, pin such achievements on our starboard @MartinSleziak
 
7:23 AM
yooooooooooooooooooooooooooooooooooooooooooo
 
meeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee
 
GUYS! How do you make $e^C$ become C
 
It is simply denoting the constant $e^C$ by another constant.
You might say that it is sloppy notation to denote it by the same letter. But this is quite often done when the actual value of the constant does not matter.
 
@MartinSleziak can you check this?
 
What you mean by this? I should warn you that I am not good in odes.
 
BTW you can write $\ln x$ instead of $ln x$.
 
oh
but still my question is ... is $e^C$ the same as a C
 
No, it is not the same. But if C is constant, then $e^C$ is also a constant. You can denote it by a different letter, if you prefer.
When I read your document starting from the line $\frac{dy}{dx} = -y$, that part seems ok to me.
Maybe if someone was to nitpicky, the woud point out that $\int \frac1y dy$ is $\ln |y|$ not $\ln y$.
 
-.-
 
Oh, now I spotted the mistake (I think).
In one step you replaced $e^{-\ln y}$ with $-y$.
But $e^{-\ln y} = (e^{\ln y})^{-1} = y^{-1} = \frac 1y$.
In the other words $$e^{-\ln y} = \frac1{e^{\ln y}} = \frac 1y.$$
I guess the thing seems a bit more readable in displaystyle.
I will have to leave. Good luck with your studying @usukidoll
 
7:35 AM
ya that was a big no no when I saw that I was like wait log laws.. that's a -1
thanks for your help
so I had $e^{lny^-1}$
which became $y^{-1}$
 
@robjohn OK. Did you write it?
Greetings
 
@Chris'ssis Not yet... we went to dinner and when we got home I went to sleep. I am taking care of a few things. I will get back to it in a while.
 
@robjohn OK, take your time. I need a clear mind before trying it again, and some pen and paper, perhaps some clear drawings.
@robjohn I think I might miss there something obvious ...
 
writelatex.com/1482027htjzrh#/3686628 how do I take the integrating factor of -xyV_z-v=0
 
8:08 AM
0
Q: PDEs with Variable Coefficents: Solve $xu_x-xyu_y-u=0$ for all $ (x,y)$

usukidollQuestion: $xu_x-xyu_y-u=0$ for all $ (x,y)$ My attempt: Our characteristic curve is in the form of $\frac{dy}{dx}$. Since our $dy = -xy$ and $dx = x$ we have the following separable equation . $\frac{dy}{dx} = \frac{-xy}{x}$ so that leaves us with $\frac{dy}{dx} = -y$ $\frac{-1}{y}dy = dx$ ...

 
I am being asked to write an essay on an aspect of differential geometry. Now, I wonder, when does a topic I cross the line. I want to talk about the Poincare Conjecture.
Any ideas?
Wait, I can just ask my Prof. xD.
 
No-one seems to answer questions in regards to optimisation theory, is this expected?
 
@Analysis same for pdes
ugh
 
8:23 AM
@Usukid, do you know anything about Variable end point optimization?
 
no do you anything about pdes
 
no
Why don't we have more friends?
 
:/
 
8:42 AM
@MartinSleziak I recently flagged 10 comments from X to Y and 10 comments from Y to X, praising each other's silly answers, lol.
Today I am targetting the silly comments of Z, lol.
I am going to watch X, Y and Z now every day and flag all their unconstructive comments, lol.
 
Don't you have anything better to do?
 
It's OK. I need some fun.
 
True dat.
 
Being here is a form of occupational therapy for me.
 
But keep it as "for fun" only.
 
8:52 AM
Yes, this site is all about fun. What else?
 
Some people take it too seriously, right?
 
I like to increase my flag count.
 
sure, np
have fun
 
I now have 34 flags, mostly from silly comments.
I think Arthur Fischer is doing a great job.
 
star the post----------------------------------------------------------------------------------------------------------------------------->
 
8:57 AM
Sorry, I don't do stars in chat.
 
np
 
9:19 AM
Maybe I should be a mod too. Then I can delete their comments myself, lol.
 
 
2 hours later…
10:59 AM
How is everyone?
 
Fine thanks, how are you?
 
That is good, I am quite good myself thank you.
So @Ice, what is your position in Mathematics education? 1st year student, 40 years of Tenure, etc
Didn't sell that very strong
 
@DisplayName how about you?
 
11:20 AM
@IceBoy 2.5th year student, enjoying mostly discrete math
 
cool, you'll fit right in here
 
@IceBoy Most people enjoy Discrete Math?
 
Hello, If $u$ est a solution such that f(u)=\min_{x\in X} f(x) is u bounded ?
 
@DisplayName yep, and all other types of math as well
 
11:23 AM
:D
 
Are you always so general?
 
perhaps
 
Nice
 
11:34 AM
is a global minimum always bounded ?
 
11:46 AM
@Chris'ssis I have showed that the estimate I was trying to prove was not possible. Now I am not sure of the estimate I need. :-(
 
@robjohn OK. I didn't start working on it yet. I need a proper frame of mine.
 
@robjohn please if i have $u'(t)=f(t,u(t)), t\ge0$ is $u(t)=-\int_t^{+\infty} f(s,u(s)) ds$ or $u(t)=\int_0^t f(s,u(s)) ds$ ? please
 
@Vrouvrou If they both converge, they differ by a constant. Thus, they are both solutions
@Vrouvrou wait... the first is the negative of a solution
 
@robjohn ok they are the same with - at the first
 
@Vrouvrou yes
 
11:59 AM
@robjohn i edited
 
@Vrouvrou oh, I see.
 
@robjohn please if $u$ is a solution and it is a global minimum so it is bounded ?
 
@Vrouvrou I edited your original statement.
@Vrouvrou in what sense is it a global minimum?
 
for a functional I, $I(u)=\min_{x\in X} I(x)$
 
@Vrouvrou what functional are you applying to $u$?
 
12:04 PM
$I$ is the functionnal associated to the problem where u is solution
@robjohn
 
@Vrouvrou I am trying to figure out what you are asking... I don't know what functional you are looking at. $u$ can be a solution to many differential equations. What functional is associated with a given differential equation?
 
I have a bvp and a solution of this bvp it the critical point of the functionnal associated to this bvp
@robjohn
 
@Vrouvrou I don't know what functional to associate with a given boundary value problem...
 
I can't seem to solve a specific proof problem in optimization theory, does anyone here have any knowledge in this field?
 
@Analysis Just ask the question on the site.
 
12:15 PM
Noone would answer because it is too boring
 
@Vrouvrou Are you minimizing a functional and deriving a differential equation? Like in Calculus of Variations?
 
@Analysis You have not even tried.
 
@WillHunting I have with two less difficult and shorter questions with no answers since the 4th
 
@Vrouvrou If $u$ is a solution to such a differential equation, it is a stationary point. It could be a maximum or minimum or neither, it need not be a global one. The same as if $f'(x_0)=0$, $x_0$; it could be a maximum or minimum or neither, it need not be a global one.
@Vrouvrou as for bounded, I don't see how that is related.
 
Huy
@rehband What's wrong with getting a girlfriend? ._.'
 
12:26 PM
@Huy Where did that come from?
 
Huy
@Analysis: Starred messages on the RHS of the chat.
 
@r9m Indeed, but I had some difficult time when learning his method
 
r9m
12:57 PM
@Huy ^^' I wish I knew =P
@Anastasiya-Romanova okay .. :) I want to get an idea of know how to choose appropriate change of variables to compute similar series .. can you help ? :-)
 
Random integrals doubt: If I have an indefinite integral like $$I = \int f(x)\,\text{d}x ,\space a<x<b $$ Does that mean I have to put in the limits of x and make it a definite integral?
I'm integrating $\cos (x) , 2n\pi - \frac{\pi}{2}<x< 2n\pi + \frac{\pi}{2}$ . So, the vagueness of this is pretty much messing with me.
 
Oh noes wait.
 
$$\sum _{n=1}^{\infty } \frac{\binom{2 (n-1)}{n-1}}{4^{n-1} n^2}=4 (1-\log (2))$$
 
@MatsGranvik: That's purdy
 
@Nick If you're just given $a < x < b$ that might also mean that you are just given the domain.
 
1:03 PM
@BalarkaSen: Yeah , it's the domain.
 
@Nick Also related to matrix inversion, sum over divisors and series expansion of 1 divided by Sqrt[1-x].
 
@Nick Then you're just doing indefinite integral.
 
@BalarkaSen: Mhh, fog lifted.
 
Definite integrals results in constants, whereas you want functions of $x$ in here.
 
@r9m Sorry, I have no idea
 
1:11 PM
@MatsGranvik Seemingly resembles this
I presume it can be derived by diffing the quadratic Lagrange-Burmann inversion expansion.
 
@BalarkaSen Yes resembles a bit.
 
I am not sure how hard it is to diff Lagrange-Burmann expansions though.
So I'll pass.
 
Anyone know any good online graphing tools?
 
Desmos, maybe?
 
I need a graph of the following: $$ y = \int{\left| \cos x\right|}\, \text{d}x =
\begin{cases}
\sin (x) + C&\mbox{, } 2n\pi - \frac{\pi}{2}<x< 2n\pi + \frac{\pi}{2} \\
-\sin(x) - C& \mbox{, } 2n\pi + \frac{\pi}{2} < x < 2n\pi + \frac{3\pi}{2}
\end{cases}
,\text {where }n \in \mathbb Z $$
Difficult to put that in
 
1:20 PM
?
 
@BalarkaSen: Is that above thing even correct?
 
I don't see how you're getting $-\sin(x)$.
 
@BalarkaSen: I just integrated the definition of the absolute value function $$\left| f(x)\right| =
\begin{cases}
f(x) &\mbox{, } f(x)>0 \\
-f(x) & \mbox{, } f(x)<0
\end{cases} $$
 
That's a false definition.
 
orly?
 
1:26 PM
oh no right
i misread.
your integral seems OK then
 
2
A: Sum of the series $\frac{1}{2\cdot 4}+\frac{1\cdot3}{2\cdot4\cdot6}+\dots$

SuperAbound$$\frac{1}{2\cdot 4}+\frac{1\cdot3}{2\cdot4\cdot6}+\cdots=\frac{1}{2}$$ Rewrite the sum as \begin{align} \frac{1}{2\cdot 4}+\frac{1\cdot3}{2\cdot4\cdot6}+\cdots &=\sum^\infty_{n=0}\frac{(2n+1)!!}{(2n+4)!!}\\ &=\sum^\infty_{n=0}\frac{(2n+1)!}{2^{2n+2}n!(n+2)!}\\ &=\sum^\infty_{n=0}\frac{1}{2n+2...

$$\frac{1}{2\cdot 4}+\frac{1\cdot3}{2\cdot4\cdot6}+\cdots =\lim_{x\to1/2} \frac{1-2x^2-\sqrt{1-4x^2}}{4 x^2}=1/2$$ Q.E.D.
bbl
 
Erm, how're you getting that limit representation @Chris'ssis?
 
1:40 PM
Nevermind, figured that this one also involved Lagrange-Burmann =P (posted as a comment in there)
 
r9m
@Anastasiya-Romanova okays ! :)
 
 
Very smart answer.
 
Wolfram alpha wins over all graph tools
@BalarkaSen: I think I like the answer via telescoping series better than the smart answer
 
1:45 PM
By "smart answer" I was referring to the telescoping, actually.
 
@BalarkaSen A masterpiece.
 
Mhh, fog lifted. Mist cleared.
 
Oh figured I linked wrong. I meant math.stackexchange.com/questions/936236/…
@Chris'ssis Isn't it? =D
Who could've though that there was such an elementary answer?
 
Low rep user, that's who!
 
@BalarkaSen Yeah. Well, playing with some terms, doing a little bit of research, one can get amazing answers.
 
r9m
1:52 PM
@BalarkaSen math.stackexchange.com/questions/919519/… ;) @robjohn the wizard :)
 
clap clap clap
 
@r9m: I will never be able to digest that identity.
 
r9m
@Nick don't try to digest an elegant spell like that at once (the trick is to get used to it first =P I guess)
 
@r9m: And to get used to it, one must first understand it.
 
r9m
@Nick hmm .. not really :P (one usually tries to get used to stuff that they can't understand completely :-) .. )
 
2:03 PM
Ok, I get it. You can drive a car without building a car.
 
r9m
^^ right
 
In mathematics, it's fatal to quote theorems without completely understanding proofs however.
Quoting theorems is journalism, NOT mathematics.
 
Random Question: If $y = sin^{-1) x$, why is $$\int_{0}^{1}{sin^{-1} x} ,\dx \equiv \int_{0}^{\pi/2} sin y \, dy \space ?$$
 
@r9m
 
Can the Hypergeometric series PFQ be used to express the Riemann Zeta function completely?
 
2:18 PM
No, @Mats
 
Too bad.
 
zeta functions are special cases of hypergeometric series, not the other way around
 
isn't that what he was asking?
 
oh right
 
@Balarka Zeta functions are special cases of hypergeometric series, but not: zeta functions are a subset of hypergeometric series?
 
2:20 PM
no i misread. nevermind.
yes, zeta functions can be expressed in terms of hypergeometric functions.
 
Jihaa!!! Yippie ja, jaa.
 
$$\zeta_H(s, a) = a^{-s} {}_{s+1}F_{s}(1, \underbrace{a, a, \cdots, a}_{\text{s copies}}; \underbrace{a+1, a+1, \cdots a+1}_{\text{s times}}; 1)$$
Set $a = 1$
 
How do I get from here:

HypergeometricPFQ[Flatten[{{1}, Table[1, {n, 1, k}]}], Table[2, {n, 1, k}], 1]

to the whole complex plane?
Is that Hurwitz zeta? @BalarkaSen
 
Yes.
In fact Lerch transcendences can also be expressed in terms of hypergeomtric functions.
 
2:36 PM
What would be the meaning of setting the first $1$ equal to $\frac{1}{2}$, like this:
$$a^{-s} {}_{s+1}F_{s}(\frac{1}{2}, \underbrace{a, a, \cdots, a}_{\text{s copies}}; \underbrace{a+1, a+1, \cdots a+1}_{\text{s times}}; 1)$$
?
That is the Dirichlet series I have in my question.
Except that the a^-s is missing in my Dirichlet Series.
 
Weird. I am not sure.
 
$$\sum _{n=1}^{\infty } \frac{\binom{2 (n-1)}{n-1}}{4^{n-1} n^s} \sim a^{-s} {}_{s+1}F_{s}(\frac{1}{2}, \underbrace{a, a, \cdots, a}_{\text{s copies}}; \underbrace{a+1, a+1, \cdots a+1}_{\text{s times}}; 1)$$
Binomials in the numerators of the sum.
 
what are the $a$s in your sum?
 
I don't know.
a=1
 
$a = 1$?
 
2:43 PM
yes
 
And is that the asymptotic notation?
 
No it should be exact. I just put the $\sim$ there to not say anything wrong.
 
@Huy It's like having a part time job. Less time for mathematics.
 
http://math.stackexchange.com/q/936433/8530
the link
 
@MatsGranvik ${}_2F_1(1/2, 1;2;1) = 2$ according to W|A, but your sum is $4(1-\log(2))$?
 
2:48 PM
@BalarkaSen I did the comparison in Mathematica:


Do[
Print[{N[{Sum[
Binomial[2 (n - 1), (n - 1)]/4^(n - 1)/n^k, {n, 1, Infinity}] -
HypergeometricPFQ[Flatten[{{1/2}, Table[1, {n, 1, k}]}],
Table[2, {n, 1, k}], 1]}],
{HypergeometricPFQ[Flatten[{{1}, Table[1, {n, 1, k}]}],
Table[2, {n, 1, k}], 1]}}], {k, 1, 4}]
....
....
The first HypergeometricPFQ[Flatten[{{1/2}, Table[1, {n, 1, k}]}],
Table[2, {n, 1, k}], 1]}],
is equal to the sum
 
Hey everyone. What symbol do I use if I want to say 'less than or equal to approximately...'?
 
It fails for $s = 2$.
 
This does not fail though:

Do[
Print[{N[{Sum[
Binomial[2 (n - 1), (n - 1)]/4^(n - 1)/n^k, {n, 1, Infinity}] -
HypergeometricPFQ[Flatten[{{1/2}, Table[1, {n, 1, k}]}],
Table[2, {n, 1, k}], 1]}]
}], {k, 1, 12}]
....
It returns zero for all integer values of $k$.
*for all positive integer values of $k$.
 
oh right i was looking at $s = 1$ in which case indeed the sum is also $2$
@MatsGranvik interesting.
try expanding the hypergeometric series.
 
@MartinSleziak In meta-meta chat, they have an entire room exactly for that -- praising moderators. chat.meta.stackexchange.com/rooms/551/in-praise-of-moderators
 
3:02 PM
@BalarkaSen Ok, here comes one expansion:

$$\text{Series}[\text{HypergeometricPFQ}[\text{Flatten}[\{\{x\},\text{Table}[1,\{n,1,2\}]\}],\text{Table}[2,\{n,1,2\}],1],\{x,0,4\}]$$
$$=$$
$$1+\left(2-\frac{\pi ^2}{6}\right) x+x^2 \left(-\frac{\pi ^2}{6}+2+\frac{\psi ^{(2)}(2)}{2}\right)+\frac{1}{90} x^3 \left(-\pi ^4-15 \pi ^2+270+45 \psi ^{(2)}(2)\right)+\frac{1}{360} x^4 \left(-4 \pi ^4-60 \pi ^2+1080+180 \psi ^{(2)}(2)+15 \psi ^{(4)}(2)\right)+O\left(x^5\right)$$
Went out of screen.
 
just do the usual expansion coming from factorial terms
 
I don't know how to do that.
 
@Thursday I am aware of that. I thought about starting a room In Praise of MSE users. But I thought that it would probably not be used too much, so I decided to post this in the main chatroom.
 
@MatsGranvik use the definition of hypergeometric functions!
 
BTW I know that the name of the room you mention has the name which looks like it could be used for messages showing appreciation for various things done by the mods, but when I peek into that room, it does not really look like that.
 
3:09 PM
@ArthurFischer Regarding this comment of yours: actually, close-votes can be retracted. Delete votes cannot, but close votes can.
Once retracted, a close-vote cannot be re-cast.
 
3:20 PM
@Chris'ssis howdy-do
 
@Hippa Herro
 
@BalarkaSen Herru
 
@Hippalectryon Always great ...
 
3:35 PM
@Chris'ssis @Hippalectryon @BalarkaSen Good evening
 
@rehband Hi
 
Haha, even after I flagged yesterday, she is still leaving stupid comments.
 
@rehband Hello. How're you doing?
 
Looks like no mod told her off. Maybe I should tell a mod to tell her to stop.
 
@BalarkaSen Quite good, and guess what: I did algebra all day today! (exam prep)
:(
 
3:38 PM
Excellent.
 
And yourself?
 
Number theory mostly.
 
Great
 
@rehband any interesting algebra problem you came upon?
 
@rehband Do you recommend any package for learning German from scratch? Something like "Teach yourself German" with a book and CD?
 
3:44 PM
@BalarkaSen There were a few that were quite nice, but I don't remember them exactly. I'm just starting to see more connections between some of these things.
 
@rehband :D
 
But a lot of it is tedious calculations unfortunatly
 
@Hippalectryon Do you recommend any package for learning French from scratch? Something like "Teach yourself French" with a book and CD?
 
@WillHunting Uh no sorry idk any :/
 
@WillHunting Sorry, I don't know any such book or CD package. My uni offers like a million language courses, perhaps yours does too?
 
3:45 PM
@rehband Well, do let me know if you come upon anything of interest.
 
@rehband Hehe, I am trying to learn by myself.
 
@BalarkaSen Will do, you're probably really good at LA too right?
 
Los Angeles? No, I live in India.
=P
 
U guys have Bollywood though, right?
 
What do you mean by LA though?
 
3:47 PM
linear algebra
 
wait, what? by "algebra" you meant linear algebra?
slowly walks away
 
Yikes
 
i thought you meant abstract algebra. group theory and whatnots.
 
Well, a little bit about that too
 
Abstract algebra is a bad term, because linear algebra is abstract too, lol.
 
3:48 PM
And a little bit of affine geometry
 
Hello @ArthurFischer thank you for your hard work.
 
@anorton Yeah, Willie Wong informed me of that somewhat after the post was deleted. Didn't feel like going back and editing the comment. Thanks, though.
@WillHunting Ummm... what hard work?
 
@ArthurFischer As a moderator, and deleting over 1000 comments, lol.
 
@rehband if you have those kind of problems (groups, etc.) then let me know.
 
@BalarkaSen Yes sir
 
3:51 PM
@BalarkaSen If you are looking for exercises, just look at more books.
 
@WillHunting math.SE isn't too hard to moderate. There are times I want to delete everything, but thankfully I can look at what mods on other SE sites have to deal with.
 
@WillHunting I am not looking for exercises.
 
Destroying the odd spam account also makes it a lot easier. ;-)
 
@ArthurFischer The past few days I flagged about 30 comments by 3 users who like to praise one another's answers. They probably upvote among themselves a lot too. I kind of detest that group, lol. I will continue to watch them.
 
@WillHunting I already have a few abstract algebra books where I can have a lot of exercises. I am looking for interesting problems to think through. Like the one Mike gave me a few weeks back.
 
3:54 PM
@BalarkaSen I see. I think that exercises and problems are the same thing, lol.
 
@WillHunting You're not getting my point. It'd take more time to dig up an interesting and hard exercise from a few thousand problems than solving it.
 
@WillHunting So that was you! Thanks. Sometimes we forget to look into their comments. But, yeah... that trio... the number of discussion we've had about that trio... [shakes head]
 
@ArthurFischer I know that you can't remove the upvotes, because they did not trigger the automatic script and can be considered "genuine", lol.
 

« first day (1506 days earlier)      last day (3506 days later) »