« first day (1939 days earlier)      last day (3098 days later) » 

6:01 PM
@robjohn It seems the value of the multivariable integral tends to $1$ while the number of variables tend to $\infty$, strictly increasing from $v\ge2$.
 
@AinzOoalGoal hey mr le pegase
 
@Chris'ssistheartist my laptop is still chugging on the negative exponent 4 variable computation.
 
@robjohn different values all over.
 
@robjohn: Hope you don't burn up your laptop the way AinzOoalGoal did.
 
@TedShifrin how did they do that?
 
6:03 PM
Ask him .... I really don't know.
 
i would smash it with a jackhammer if he wouldnt respond after an hour or so
 
Salut, @Agawa.
 
salut , ça va bien bien Mr ted ?
 
Oui, bien, merci, et toi?
 
On fait rouler Mr ted
 
6:18 PM
@Agawa001 hey
 
What are you studying these days, M le Méchant?
 
@TedShifrin Probabilities
 
Unlikely circumstance, but ok :)
 
I also tried to see that integral using differential equations, that is
$$y''+y=\Gamma(1+x)$$
 
@Chris'ssistheartist Inverting that differential operator is not a difficult process
 
6:33 PM
I considered the case $a=1$.
 
@Chris'ssistheartist $a$?
 
2
Q: Evaluating $\int_0^{\infty} \frac {e^{-x}}{a^2 + \log^2 x}\, \mathrm d x$

LTSI am trying to evaluate this integral: $$\int_0^{\infty} \frac {e^{-x}}{a^2 + \log^2 x}\, \mathrm d x$$ for $a \in \mathbb R$. (of course with $a > 0$) Any ideas?

 
Ah
 
I considered $$y(s)= \int_0^{\infty} \frac {x^s e^{-x}}{a^2 + \log^2 x}\, \mathrm d x$$ where above I chose $a=1$.
 
@Chris'ssistheartist Yeah, I guess that from the differential equation
 
6:35 PM
@robjohn It's a nice differential equation. I wonder if one can get a nice form. My Mathematica isn't able to do it.
 
@Chris'ssistheartist It is easy to invert if you can integrate the function against $e^{ax}$ for complex $a$.
 
@robjohn Maybe I should consider the differential equation while taking the limit from $s$ to $0$, and hope not to say a stupidity.
I'm interested in that function near $0$.
 
How can I show that P(B) = 1 entails P(A) = P(A ^ B) from the probability axioms?
 
@robjohn maybe you can do it with some complex analysis. It looks like those questions that ask for complex analysis.
 
6:57 PM
@Chris'ssistheartist I don't know. I may try it later.
@Chris'ssistheartist @r9m Doh! of course it works for both of those integrals... for any real $a$, we have $$\int_0^1\int_0^1(xy)^{axy}\,\mathrm{d}x\,\mathrm{d}y =\int_0^1x^{ax}\,\mathrm{d}x$$
that includes $a=\pm1$
 
@robjohn Nice.
 
@Chris'ssistheartist $a=0$ requires a division by $0$ in my proof, but that case is trivial.
 
@robjohn You might ask people to prove that $$\int _0^1\int _0^1\frac{1-(x y)^{x y}}{x y \log (x y)}\ dx \ dy=\int_0^1 \frac{1-x^x}{x \log (x)} \, dx$$
@robjohn You might propose something like that to a magazine.
The one above looks gorgeous.
It looks nicer after putting 1 in front.
 
@Chris'ssistheartist As long as the integrand is a function of $x\log(x)$.
 
@robjohn I integrated from 0 to 1 with respect to a.
 
7:04 PM
@Chris'ssistheartist did it work?
 
@robjohn Sure! Numerically is perfect.
 
@Chris'ssistheartist which integral are you talking about?
 
@robjohn $$\int_0^1\int_0^1(xy)^{axy}\,\mathrm{d}x\,\mathrm{d}y =\int_0^1x^{ax}\,\mathrm{d}x$$
@robjohn integrate both sides with respect to $a$, from $0$ to $1$.
 
@Chris'ssistheartist Oh, yeah. It is true for all real $a$
 
@robjohn Many other integrals can be got from exploiting that result. It's very nice, indeed.
@robjohn Just look at this one ...
$$\int _0^1\int _0^1e^{\large (x y)^{x y}}dydx=\int_0^1 e^{\large x^x} \, dx$$
 
7:09 PM
@Chris'ssistheartist If $f(x)=g(x\log(x))$, then $$\int_0^1\int_0^1f(xy)\,\mathrm{d}x\,\mathrm{d}y =\int_0^1f(x)\,\mathrm{d}x$$
 
@robjohn Yeap.
@robjohn above I replaced a by k, then I divided both sides by $k!$ and performed the summation from $k=0$ to $\infty$.
 
@Chris'ssistheartist yeah, but $e^{x^x}=e^{e^{x\log(x)}}$ so it fits my criterion
 
@robjohn If I'm not wrong this relation can be used as a recurrence relation to get multiple integrals, long crazy integrals.
Just treat the inner integral in the left side considering the initial relation, that is $\int_0^1\int_0^1f(xy)\,\mathrm{d}x\,\mathrm{d}y =\int_0^1f(x)\,\mathrm{d}x$.
Something like $$\int_0^1\int_0^1f(xyz)\,\mathrm{d}x\,\mathrm{d}z =\int_0^1f(x y)\,\mathrm{d}x$$
and then the process can continue indefinitely
 
@Chris'ssistheartist if $f(x)$ meets my criterion, it is not necessary that $f(ax)$ does. Otherwise, your equality above would give the three variable equation.
 
@robjohn Yes, I see that.
@robjohn That works using the condition you stated above, of course.
 
7:22 PM
@Chris'ssistheartist No, if $f(x)=g(x\log(x))$ then it is not necessary that $f(ax)=h(x\log(x))$
 
@robjohn Right. I only referred to the way I got my relation above with $f(x y z)$.
 
Hello hello.
 
@robjohn Yes, right, it doesn't work always.
 
So, uh, given function $f(x) = \frac{1}{xln(x)ln(ln(x))}$, I'm trying to find a concise way to list the domain.
I know it can't be negative, or zero, or $e$.
or $1$.
 
@Owatch to which thread is that directed?
 
7:30 PM
But I can't say it's $(1, +\infty)$.
What do you mean Robojohn?
It's not directed at any particular thread. It's a question.
 
@Owatch to whom are you talking? (this is why I like using the arrows showing to what you are replying)
@Owatch Oh... I see.
 
@robjohn To everyone. I don't like to ping since it sometimes irritates people. But I'll d it to reply.
 
@Owatch it would need to be $x\gt 1$
If $x\lt 1$, then $\log(\log(x))$ would be imaginary
@Owatch or are you allowing complex values?
 
If $x$ is $e$, then $ln(e)$ is 1, which is also imaginary.
Well, it's not, but because it's enclosed within another $ln()$, it becomes so.
@robjohn No, I am not allowing them.
 
@Owatch then you would want $x\gt1$
 
7:36 PM
$e > 1$, and is forbidden. Are you sure that suffices?
 
@Owatch I am just looking at $\log(\log(x))$. That is real if $x\gt1$
Since it is in the denominator, $x\ne e$
 
Why can't x = $e$, if it is in the denominator?
It doesn't work for this function, but you said the domain is $x > 1$.
which I thought $e$ would fall into.
 
user174558
@Owatch Your LaTeX is terrible, should be $x=e$, lol.
 
@JasperLoy How is it going today?
 
You just waited till my 3 minutes were up, then told me -_-. I could have edited it.
 
user174558
7:43 PM
@TedShifrin I am sorry if I upset you, but I wasn't really specifically referring to you that day. Happy Thanksgiving. You are a great mathematician.
 
user174558
@Chris'ssistheartist So so. I have yet to look through my book.
 
user174558
@Owatch Owatch sounds familiar, I have seen you before in here...
 
@DanRust @TedShifrin Let $w$ be the matrix with 1s on the diagonal, 1 in the upper right corner, and 0s elsewhere. Then $H_3(\Bbb Z)\times\frac{1}{n}\Bbb Z$ mod $\langle (w,-1)\rangle$ is torsionfree, has finite presentation $$\left\langle w,x,y,z~\left|~\begin{array}{c} [x,y]=z=w^2 \\ [x,z],[y,z],
\\ [w,x],[w,y],[w,z] \end{array}\right.\right\rangle$$ and abelianization $\Bbb Z^2\times\Bbb Z/n\Bbb Z$.
 
r9m
@Chris'ssistheartist really nice idea! :D .. (I still haven't computed the infinite series you showed me in chat the the other day ... I am lazy (level: Sloth ) :P)
 
@JasperLoy You know it.
You have.
 
user174558
7:46 PM
@r9m I see your new picture.
 
user174558
@Owatch Yes, you are The Oracle or something.
 
r9m
@JasperLoy :-) One-Punch man (anime)
 
Wait, what
 
@r9m @robjohn still there is a thing that bothers me, that OP asks teh following question that is a particular case of this problem
19
Q: Prove $\int_{0}^\infty \frac{1}{\Gamma(x)}\, \mathrm{d}x = e + \int_0^\infty \frac{e^{-x}}{\pi^2 + \ln^2 x}\, \mathrm{d}x$

LTSI came across this nice identity: $$\int_{0}^\infty \frac{1}{\Gamma(x)}\, \mathrm{d}x = e + \int_0^\infty \frac{e^{-x}}{\pi^2 + \ln^2 x}\, \mathrm{d}x$$ Is there an elementary proof?

On wikipedia I saw that integral in the left-hand side, not sure where, and it doesn't have a known closed form. Now, to find the generalization for that integral in the right-hand side, given this context, it seem like solving an open problem.
 
@DanRust @TedShifrin I meant =w^n in the presentation.
 
7:49 PM
@Chris'ssistheartist so that bodes ill for the other integral :-)
 
@robjohn lol, it seem so.
The constant given by that integral even has a name ... (let me remember)
 
r9m
@Chris'ssistheartist !! Nice Identity !! .. seems OP is trying to reach a generalization of some kind!
 
So, are we all sure that the domain is actually $x > 1$ ?
I guess it works. !
 
r9m
@Chris'ssistheartist Nice there is a link in the comment under the question!
 
But I thought $e$ would be required in it.
 
7:52 PM
Fransén–Robinson constant
The Fransén–Robinson constant, sometimes denoted F, is the mathematical constant that represents the area between the graph of the reciprocal Gamma function, 1/Γ(x), and the positive x axis. That is, The Fransén–Robinson constant has numerical value F = 2.8077702420285... (sequence A058655 in OEIS), with the continued fraction representation [2; 1, 4, 4, 1, 18, 5, 1, 3, 4, 1, 5, 3, 6, ...] (sequence A046943 in OEIS). Its proximity to Euler's number e = 2.71828... follows from the fact that the integral can be approximated by the standard series for e. The difference is given by and also by ...
@r9m @robjohn ^^^
OK, one can say that a closed form can be got for that specific value of $a$, that is $a=\pi^2$ :-)
Thus
$$\int_0^\infty \frac{e^{-x}}{\pi^2 + \ln^2 x}\, \mathrm{d}x=F-e.$$ where F is Fransén–Robinson constant.
I later saw that one user specified it in the comments. I was aware of this constant since a long time ago.
 
hello
 
Overall, finding some nice infinite series for expressing such integrals is at least as interesting and nice as finding a closed form. Ramanujan did it again and again, read his notebooks.
 
r9m
@Chris'ssistheartist cool!! Seeing this for the first time here .. :)
 
@Chris'ssistheartist Yeah, using that constant that is pretty much tied to the integral. Mathematical juggling
 
Do you know where a complete beginner could learn about these amazing integrals and sums?
 
r9m
8:01 PM
@Chris'ssistheartist it's there Ramy's notebook?
 
@r9m hehe. When I looked through the OP's questions I started from this idea that he met a similar integral to the posed one, and I was right ...
 
r9m
@Chris'ssistheartist that was sharp of you! :D (saved my rest of the night .. )
 
@Brennan.Tobias MSE is a good start. :-)
@r9m :D
@r9m Ramy's notebook? It's like the bible for someone that loves the area of integrals, series and limits. :-)
 
r9m
@Chris'ssistheartist I mean the integral .. did you mean it's there somewhere in his notebook?
 
@r9m I don't remember I saw it there.
 
r9m
8:06 PM
@Chris'ssistheartist 'kay .. I'll dig through it after my exams then .. :-)
 
@r9m I only said that Ramanujan used to express ugly integrals using very nice series (it mattered less our present expectations about closed-form).
@r9m Wish you success then. :-)
 
r9m
@Chris'ssistheartist ah! I see .. :)
@Chris'ssistheartist I ain't gonna work with it anymore .. the appearance of a new constant out of the blue was enough for me :P
 
OK, put a name to the constant you get and you have a new closed form (in general, when one is tempted to say there is no closed form). Case closed.
 
r9m
lol .. didn't mean it like that :P (I had enough for a night .. that's what I meant)
 
@r9m :D
@r9m Do you want me to say a funny thing? :-) Well, related to the story with my series, I was thinking you got the closed form, but you wanna get a nice answer, and when I ask you how you did it you then you tell me you did it nicely without Mathematica stuff. :D
 
r9m
8:12 PM
@Chris'ssistheartist sure .. I like funny things :P (even at the cost of self-humiliation :P)
 
can someone make a notebook ? i mean a not-a-book :p ?
 
@r9m lol, not self-humiliation :D
@r9m Sorry for my English above, I might have put it better. :-)
@r9m that series has a special evil smile! Every time I think of it I'm about to laugh a bit. :-)
@r9m seriously speaking, people pay a lot of money to watch comedy movies, to get that entertaining stuff, but I don't remember a moment when I had more fun than in the period I did a lot of mathematics!
Never try to imagine me being that kind of serious face doing math, I'm very serious about doing math, but I have a lot of fun, mathematics makes me laugh a lot! :-)
@Brennan.Tobias Also try Paul Nahin's book, Inside Interesting Integrals, you'll have a good time with that book.
 
cool :)
 
@Brennan.Tobias It's not about rigourness, but about techniques in integrations, and the book is very friendly with new people in integration, not the could language of mathematics that might give you creeps.
 
sounds really fun I'll try it out
 
8:27 PM
@Brennan.Tobias You'll be very relaxed with that, no stress at all. I personally like very much the idea of writing such a book.
Never hide the entertaining part of mathematics.
 
i downloaded that notebook because of magic square
really nice
 
Yeah
@Brennan.Tobias Then, you can try Ovidiu Furdui's book. This one is also very nice but you need to know some stuff.
 
thank you
 
@Brennan.Tobias you can learn a lot, not only integrals, but also nice series and limits. It treats more the integrals with fractional part, but this is a class of integrals pretty cool.
 
r9m
@Chris'ssistheartist perfect math-head you are :P :D
 
8:36 PM
@r9m :D
@Brennan.Tobias I would recommend that you try to come up with your own solutions after a while, if possible, for each problem in these books. It's a good way to learn stuff.
@Brennan.Tobias also see the last one
 
oh fantastic :)
 
@Chris'ssistheartist And soon there will be your book there too :-)
 
@Brennan.Tobias Victor Moll also published another 2 books I don't have, they are more in the spirit of some work they previously published in some papers.
@AinzOoalGoal Hope so. :-)
 
user174558
8:52 PM
@Chris'ssistheartist As irresistible as Monica.
 
i just use amazon to pick on books i torrent-request em afterwards
 
@Brennan.Tobias I find those papers a very good stuff to study although at first sight one might say Ah, nothing very special. There is much power in those papers. They remind you of some very useful known formulae plus some modified forms for some integrals (again, known).
 
user174558
@Agawa001 Torrent? Have you heard of Genesis Library?
 
@JasperLoy :D
 
@Chris'ssistheartist I edited those comments so that they don't take as much room.
 
user174558
8:54 PM
@Sajindia What do you wanna talk to me about? I hope you are feeling better.
 
@robjohn Thank you.
@robjohn I should have used [text](link) - if this still works.
 
most of links are dead there :(, nice place thu
 
user174558
@robjohn I am not complaining, but this is a peculiarity of this room. In other rooms, these things are encouraged to expand, hence the feature in the first place.
 
i would torrent chris's book too :p
 
@Agawa001 :D
 
user174558
8:55 PM
@Agawa001 Have you tried libgen.io
 
@Agawa001 You don't find me yet on any torrent. :-)))))
 
user174558
@Agawa001 I cannot find torrents for most books.
 
user174558
@Chris'ssistheartist Can we also torrent Monica?
 
wow last one was very helpful, to my fav immediatly :D
 
@JasperLoy hahahaha, that was good! :-)))))
 
user174558
8:57 PM
@Chris'ssistheartist I talked to myself for 30 minutes just now. I feel better now.
 
user174558
But I sent Jonas two emails and he has not replied. Sad panda...
 
@JasperLoy Nice. Maybe I should do the same, but the problem is that I prefer to think of some math more than spending time talking to myself (especially in this period of time). :D
 
user174558
@Chris'ssistheartist Actually, I talk to myself multiple times a day. It's my own psychotherapy.
 
@Chris'ssistheartist if u have some of these clients in ur laptop, close em immediatly, they ar a perfect egress for system-penetrators
 
@Agawa001 I don't have. Anyway, while on internet you're not safe as a general rule, if you have some important data to protect.
 
user174558
8:59 PM
@Chris'ssistheartist @robjohn You might as well edit this link as well then. =)
 
@JasperLoy there are a lot of extended conversations with rendered mathjax. It is best for those to be able to look back as far as one can.
 
@JasperLoy I heard it's a powerful tool for healing.
 
user174558
@Chris'ssistheartist If I did not talk to myself, I would be dead by now.
 
@JasperLoy If there are a number of expanded links like that together, it reduces the amount one can see back without having to scroll.
 
@JasperLoy I believe you.
 
9:02 PM
@JasperLoy are you an only child? (no siblings)
 
user174558
The book I got in my mailbox today is 'A first course in mathematical logic and set theory' by O'Leary. I will start reading it on New Year's Day.
 
user174558
@robjohn Yes, but in this case, even if one has people to talk to, one should still talk to oneself.
 
user174558
It is a beautifully typeset book with two little birds on the cover.
 
@JasperLoy Only children often talk to themselves.
 
user174558
@robjohn Or mentally ill people.
 
user174558
9:04 PM
By the way, for those interested in Zorich's Mathematical Analysis, a new edition will be out soon.
 
user174558
And for those interested in Gratzer's More Math into LaTeX, a new edition will be out soon too.
 
@JasperLoy I bet the mentally ill people are not talking to themselves, but to perceived others.
 
user174558
@robjohn Aha, that is the schizophrenic people.
 
user174558
Munkres's Topology is very expensive. One should get Willard's General Topology instead for a thorough treatment of general topology.
 
user174558
I have no idea why Munkres is so popular, other than it being a classic.
 
r9m
9:09 PM
@robjohn plenty of examples among grown ups too .. some people can't leave that habit as they grow up .. (just like reading by pronouncing every word in mind .. )
 
user174558
@r9m Pronouncing
 
user174558
@r9m Also, one dot is for full stop and three for ellipsis, no such thing as two dots.
 
r9m
@JasperLoy thanks .. !
 
user174558
I think two dots arose online or in SMS.
 
r9m
@JasperLoy . .. ... .... ..... ...... ... :P
 
user174558
9:12 PM
Lang likes to reference Greenberg's algebraic topology book. I think because he is his student.
 
r9m
@JasperLoy who is Monica?
 
user174558
@r9m Monica Bellucci is an Italian actress.
 
@r9m A cute girl. :-)
 
r9m
@JasperLoy ah!! :D
 
user174558
@r9m She likes Monica, but I like Laura Ramsey.
 
9:14 PM
@JasperLoy I found munkres inscrutable. needs more pictures.
 
user174558
@Mitch I think it already has many pictures.
 
Oh.
 
user174558
@Mitch I prefer no pictures at all, because I draw them myself.
 
Not enough.
for me
 
user174558
I think pictures should only be there if words are not clear.
 
r9m
9:16 PM
@Chris'ssistheartist Have you seen this recent article by Furdui? :-) We dealt with similar problems in past I think :-)
 
user174558
A word is worth a thousand pictures.
 
Not to be antisolipsistic, but when you read topology, what is really going through your head is pictures.
 
user174558
Yes, but a word is precise while a picture is not.
 
user174558
Pedagogically pictures are important, but one should draw them oneself.
 
but the picture communicates things more quickly
 
9:17 PM
@r9m No, I didn't.
 
user174558
@Mitch Yes, but it is against my taste.
 
I don't think you taste very good either.
 
user174558
Yes, you are right.
 
ha ha
 
user174558
Wait, what did you mean?
 
user174558
9:18 PM
Are you a cannibal?
 
actually with algebra... yes, there arent exactly pictures going on but... but there is something visualized.
 
r9m
@Chris'ssistheartist 'kay :-) I kinda remember doing the first problem .. (in a discussion in chat)
 
it's not just pushing symbols around.
 
@r9m Yes. Do you remember when? We did more series with $\psi$.
 
r9m
@Chris'ssistheartist I blogged about it back in 2014, so must be around then. Lemme check.
 
9:20 PM
in English Language & Usage, Sep 15 at 14:09, by Mitch
Cannibalism jokes are funny because eating people is wrong
bbl
 
user174558
@Mitch I think in history there were people who ate dead people to survive.
 
r9m
@Chris'ssistheartist @Robjohn awesome solution
 
@r9m But aren't these problems different?
 
r9m
@Chris'ssistheartist yes .. but the basic principle was the same .. we were doing Able Summations :)
 
9:26 PM
@r9m Yes. Then there is another thing: he also used the Flajolet's paper. Initially I thought he avoided it. Did I tell you I calculated all series in Flajolet (I know the statement seems crazy, but I did it)?
@r9m I also have for my book such versions.
 
r9m
@Chris'ssistheartist yes you did, I remember! :D And I gasped with disbelief :P
 
@r9m I had no idea of this paper.
@r9m I know, it's hard to believe it. :-) Just wait for my book.
 
r9m
@Chris'ssistheartist neither did I until today noon :)
 
@r9m I know. It's so crazy to believe it. I don't condemn you at all, you're right to show disbelief, I might also do it if I were you.:-)
 
r9m
@Chris'ssistheartist I'll brush my teeth with your book :P (once I get ahold of it that is .. )
 
9:31 PM
@r9m lol, that's a statement!!! :-))))))
@r9m One thing: all I say about mathematics is saint. :-) If I say anything about math, it's exactly like that.
Let's wait then. :-)
 
r9m
@Chris'ssistheartist I am waiting patiently :)
 
@r9m how you got that paper, btw?
@r9m I got it.
 
r9m
@Chris'ssistheartist researchgate (I am stalking Furdui as well :P)
 
@r9m When I said "only children", I didn't mean only children :-) I still talk to myself when thinking deeply.
 
9:44 PM
@r9m How did you join it ? Do you have an institution email ?
 
r9m
@robjohn :D
 
@r9m Just saying that sometimes life surprises you so nicely.
 
r9m
@AinzOoalGoal yas
 
lucky
 
r9m
@Chris'ssistheartist :-)
 
9:47 PM
Anyway. I learned many important lessons in the last period of time.
 
r9m
@Chris'ssistheartist who is not Romanian?
@Chris'ssistheartist who? who? who? :D
 
@r9m Not a known person probably for you. Anyway, it's premature to say anything about it.
 
:25697492 do you have a publisher?
 
@robjohn Yes.
 
@Chris'ssistheartist great! Have they given a deadline?
 
9:51 PM
@robjohn There is a deadline too.
The way it is.
Anyway, I learned many tough lessons (that almost drive you completely crazy), and I know exactly what to do in the future in all my mathematical activity.
My book is the proper response to everything, let its stuff quality define me.
 
r9m
@Chris'ssistheartist :D
:25697752 well .. pain builds character :-) (that's what I keep in mind when I face hard situations :-) )
 
Anyway. I don't say anything else.
@r9m Good.
 

« first day (1939 days earlier)      last day (3098 days later) »