« first day (796 days earlier)      last day (4230 days later) » 
00:00 - 18:0018:00 - 00:00

user19161
12:37 AM
@OldJohn He is an unregistered user there, not deleted.
 
12:52 AM
@skullpatrol how is your sunday?
Do you think is possible to build a tricorder?
@infinitesimal these questions drive ME CRAZY!
 
OK so far, bye for now...
 
:(
 
;-)
 
:'
@skullpatrol do yo think is possible to build a tricorder?
 
1:12 AM
@Charlie I'm not sure, sorry but I must run ..
 
@skullpatrol oh..please..go
sorry
 
bye
 
bye
 
user19161
1:28 AM
@Charlie Hey there.
 
A wild Jasper Loy appears...
@JasperLoy sup?
 
user19161
1:53 AM
@N3buchadnezzar Haha. How is your studying coming along?
 
@JasperLoy It is going mediocre, guess what time it is here in norway =)
 
user19161
@N3buchadnezzar Ah, you should be sleeping then!
 
4 am, soon time for dinner
 
J G
4am is dinner time!
is that like dinner number 3?
 
first =)
Proving that $\text{Im}(z^2)/|z|$ is continous is hard...
Need to show that $\lim_{z\to0}\text{Im}(z^2)/|z|=0$ by using the $\epsilon/\delta$ definition.
 
2:16 AM
People are dropping like flies
two in the last three days
 
J G
where are they dropping?
i'm here!
 
 
1 hour later…
3:41 AM
@robjohn Rob.
 
@PeterTamaroff yes?
 
@robjohn I'm stuck.
@robjohn Do you have strong math-arms?
 
@PeterTamaroff oil or WD-40?
@PeterTamaroff just semi-automatic...
 
@robjohn WD-40 seems OK.
@robjohn Its the Riemann Lebesgue Lemma.
 
3:44 AM
@PeterTamaroff what's the problem?
 
I have proven before this that $f$ is integrable iff for every $\epsilon >0$ there exist step functions with $s_1\leq f\leq s_2$ such that $$\eqalign{
& \int\limits_a^b f - \int\limits_a^b {{s_1}} < \epsilon \cr
& \int\limits_a^b {{s_2}} - \int\limits_a^b f < \epsilon \cr} $$
@robjohn Now, I proved the RL lemma for step functions
So I should use that above to prove it for any integrable $f$
But I can't get to bound $$\int\limits_a^b f \left( x \right)\sin \left( {\lambda x} \right)dx$$ between the integral of step fns and $\sin \left( {\lambda x} \right)$
It is probably something really stupid, but I'm missing it.
I'd prefer you hinted me.
 
4:02 AM
@PeterTamaroff If you know fourier series the riemann lebesgue lemma is easy
 
4:13 AM
@PeterTamaroff First $\int\limits_a^b s_1 \left( x \right)\sin \left( {\lambda x} \right)dx$ goes to $0$ as $\lambda$ goes to infinity,late use the fact that any $f$ can be approximated with step functions, the RL follows for $f$. I had read such a proof but I am not quickly able to recollect it.
 
 
1 hour later…
5:27 AM
@JasperLoy This is sad
 
That is excellent inspiration for next April.
 
@anon =) Hey
@anon How's it going?
 
meh
 
5:48 AM
@anon meh bad?
 
meh is neutral or feelinglessness
 
@anon Heh well, yeah, nevermind.
@anon Did you see my RL problemup there?
 
dunno much about that stuff
 
@anon I guess you're more into algebra, right?
 
I never looked much at the tedious stuff in analysis
 
5:54 AM
@anon Why "tedious"?
 
Because it involves details that bore me.
 
@anon Hehehe maybe! But one gets used to those details.
@anon What are you studing at the time?
 
I want to get my hands dirty with some computations and look at structure constants in representation rings, and then see if I can graduate to finding ways to describe schur functors' effects. (I have found schur functosr to be a spiritual huge generalization of the newton-girard identities for symmetric polynomials via taking traces, which is what led me to them more or less.)
so IOW representation theory
 
@anon IOW?
@anon Ok, whatever those are!
@anon Is the fundamental theorem of symmetric polynomials complicated to prove?
 
@PeterTamaroff No, it's very quick. I've paraphrased it here and here for instance.
Do you know about the NG identities at all?
 
6:04 AM
@anon Well, now I do...
@anon (I like to explain math to people that don't know about it, sometimes)
 
The formulas in sections 2.2 and 2.4 on the NG WP page correspond to traces of the results of schur functors for the alternating and trivial representations of the symmetric groups $S_n$ respectively.
 
@anon What is a functor?
 
@anon Oooook. I'll sleep now.
@anon Bye byes.,
 
@PeterTamaroff Roughly speaking, categories are comprised of mathematical objects and certain types of maps between them; a functor maps objects from one category into another (as well as maps, aka morphisms or arrows, from one into the other category) in such a way information like $A\xrightarrow{f} B$ entail corresponding $FA\xrightarrow{Ff}FB$.
Later.
 
6:38 AM
Hi
 
 
3 hours later…
9:41 AM
Why are all these people leaving?
 
@robjohn who all? I got a -10 today from a deletion of a user.
 
@JayeshBadwaik 3 have left in the last 3 days.
I lost 134 yesterday
 
@robjohn hmm. Probably SE network can have a "Why are you leaving?" question to those who leave? This might help.
 
@JayeshBadwaik For those who leave in the future. It is too late for those who've left already.
 
@robjohn Yup. Some people can leave for no reason at all, like Jasper left Ubuntu.SE and wants his online presence clean.
 
9:48 AM
@JayeshBadwaik clean? what does that mean?
 
@robjohn They do not want to have accounts if they are not active there. This way, they don't have accounts of which they have forgotten about.
 
@JayeshBadwaik That's lame. If they're not planning on going back, what does it matter if they forget about it?
I'd just make a note and file it away for later reference, if needed.
 
@robjohn I know. But that is what people have in mind. One of my friends told me he was worried about being hacked without being known. Probably they are a little careless with their info online and hence, are afraid what info of theirs can be used if some of the accounts are hacked. I don't know, but this might be a reason. LinkedIn and Sony Hacks do not provide any reassurance.
 
There is very little personal information recorded here. An email address and some IP addresses. If they want to go they can even remove the email address and their "about me" info
 
10:03 AM
Holy cow, it is like 12 already!
 
@robjohn Did you mean LVK? He was very productive for the short time he spent here.
 
@MartinSleziak He took 134 of my points with him. I don't know how he did that...
 
I wonder what will happen to the bounties he awarded. (He has put 10k+ rep into bounties, basically all that he earned.)
I've read the post at meta.SO, but various people seem to give various answers there: Bounties from deleted accounts should still be awarded; al least, the indicator should be removed.
 
@MartinSleziak bounties are points that were given up by someone and cannot be gotten back by them, so they should not leave with them. However, I guess they do, is that right?
 
Grace Note wrote there: Deleted users, like deleted posts, revert all reputation changes that they influenced.
Although active bounties (not yet awarded) are taken by the Community user, according to that post.
Anyway, the post is from 2011, things may have changed.
 
10:11 AM
@MartinSleziak Yes, but there has been discussion of having accounts not really deleted, but held almost intact, at least for reputation, etc.
 
Reputation issues aside, I wish he stayed...
 
LVK. Why don't I know him?
 
a.k.a. Leonid Kovalev
 
@MartinSleziak I wish these people all stayed. Except for a few trolls, most people here contribute very good stuff.
 
@JonasTeuwen Here is google cache of his page, you can see which tags he contributed most.
 
10:16 AM
Oh, Leonid!
What are the reasons for leaving?
 
I have to go, see you later...
 
Bye!
 
@MartinSleziak later
@JonasTeuwen anything interesting on schedule for today?
 
No. Write a paper 8-(.
I never get any done. 8-).
I always find something cool (for me) when I try to write that.
The annoying thing somebody else probably already figured out.
Like rigged Hilbert spaces.
I called them Nuclear Sandwich Spaces, but they are... the same 8-(.
 
@JonasTeuwen rigged?
 
10:21 AM
Yes, you add distributions to a Hilbert space.
So have some unbounded operator $L$ and define a norm $\|u\|_n = \|L^n u\|$. Then for the Schwartz functions as domain the intersection of all the Hilbert spaces $\mathcal H_n$ are Schwartz again and dense in $L^2$, their union are the distributions.
So you sandwich $L^2$ between those.
Allows you to just be like "whatever" and manipulate distributions like functions 8-).
Also eigenfunction decompositions become true for generalized functions ánd, the decomposition is just an integral now.
I just wanted to use the distributions "informally", but as mathematicians no like that, then I thought: oh well! Let's add it to the space!
But some of these crazy quantum dudes have done that already. And many others.
Kinda $L^2$ on steroids.
 
physicists like to be able to cite real math for their handwaving.
 
I believe a real mathematician made that construction 8-).
But it kind of makes sense. Kind of.
 
and this allows them to use distributions in a freely hand-wavable way.
 
Yes!
Well, it makes the hand-wavable way fine.
But you should still know what you are doing.
 
@JonasTeuwen Oh, yes. Mathematicians do the physicists' dirty work
 
10:28 AM
I was more interested in having spectral decompositions "similar".
@robjohn Yes, but I believe that physicists do not really care if they are not certain their theory is the good one.
It is like you would make some calculations and do a bit hand-wavy in the middle. If it works out you go back and figure out why you can do that.
But I thought it was kinda cool as it gives a quite explicit construction of both spaces in terms of the operator you would care about.
But it is quite frustrating that in 9/10 of the cases when I find something cool somebody else did that already! And the 1/10 is less cool, but okay.
But, nice. So if it is constant you cut out some polynomials in each step. No polynomials allowed!
 
10:57 AM
@robjohn And engineers like to cite physicists! For example, there is a paper published in IEEE which proves that the most efficient way to charge a capacitor is by keeping the charging voltage marginally higher than the current voltage of the capacitor!
 
11:40 AM
hi @JasperLoy
hi @anon
 
user19161
@skullpatrol Hello. Here comes the cool anon too.
 
holey moley
I'm just checking if there's a quiz scheduled for class this morning. if not, I will go back to sleep because I have a headache
 
which class?
 
user19161
@anon Oh dear, you should not work too hard. And get more sleep. And spend less time here. Take care.
 
user19161
@skullpatrol SKULL's class on "minus versus negative".
 
11:45 AM
@JasperLoy You need to put up the maximum bounty of 400 reputation points to enter that class.
 
user19161
@skullpatrol It's OK, I am only in there to take out the trash.
 
12:59 PM
Hi @robjohn : I have the pdf and tex file (isometric embeddings problem discussed last week) ready. Waiting for you to come online to send it as file transfer.
 
 
2 hours later…
2:57 PM
@robjohn Could you take a quick look at my comment/answer here?
please...
 
@infinitesimal sounds kind of contrived, and why a linear combination? that leaves the coefficient of the argument up in the air.
 
@robjohn That's all I could come up with. With the solution of a indefinite integral there is an added arbritrary constant added, what is the name for that combination?
 
@infinitesimal there is no similar concept, because the idea of a "multiple" of something has to do with integer divisibility and there is no such concept for addition.
 
@JayeshBadwaik I really need emacs bindings for Mendeley! This stuff sucks donkeys from hell!
@robjohn My advisor thought my nuclear sandwich space was getting out heavy machinery to do something simple 8-(.
"I work with heavy machinery, bro" will hence be the title of that chapter.
 
3:13 PM
@JonasTeuwen what is it you are trying to do?
 
@robjohn I am not sure.
Do things with operators $e^{tL}$ where $L$ is an unbounded operator.
 
@JonasTeuwen Ah, therefore you use the dynamite and see what it shakes loose?
 
@robjohn Yes!
Seems to be OK to me.
@robjohn Basically what they do is the same: define it on a set where it makes sense (basis), then do whatever and use density.
But I just make sure I can shoot it all, and when it is all dead, I check if it was what I wanted to kill. Just a change of perspective 8-).
 
@JonasTeuwen I have an answer that I expanded on over the weekend. Would you look at it and see if it looks clear to you? I am too involved to tell.
5
 
Sure.
@robjohn Nice!!!
So if I understood correctly, you have a contour. You figure it fails at many points and you compute the contours between and the ones through the singularities?
 
3:21 PM
@JonasTeuwen so it was not too involved and got the point across?
 
If that is the point, then it got it across.
 
@JonasTeuwen yep
 
It is the only non-standard idea for computation.
The rest is being smart.
And have some stamina...
 
@JonasTeuwen yeah, it is a standard contour integral, except that there are many singularities and you need to miss them and show that the integral along the one contour vanishes as $R\to\infty$.
 
Yes, got it.
But. Not sure if I am the right person to judge, I might know too much?
 
3:24 PM
@JonasTeuwen cool, then I will leave it alone :-)
@JonasTeuwen I will ask someone else, but they have to know about contour integration
@JonasTeuwen: the other thing that is a bit confusing is that the two sets of singularities almost lie on top of one another.
 
I need two thesis.tex versions. One that is advisor friendly. Perhaps I need a git branch "advisor".
@robjohn Yes. But I think you gave enough information such that if they do not understand and would be able to they would figure out.
A bit like Stein's books. 8-).
 
@JonasTeuwen With comments.sty it is possible to compile two versions from the same file. I use this package often. See tex-faq.
 
Excellent!
 
Or more then two versions, if needed.
BTW we have another 100k+ user.
2
 
@MartinSleziak Taking over for Arturo in that spot.
 
3:33 PM
hello
 
@MartinSleziak The coolest would be if you could mark pieces and they are automatically branched in a git repository...
 
I do not know git, sorry....
 
There is also versions.
Git is just versioning management for software.
 
@JonasTeuwen You can use "\if" and command line arguments defined in a makefile to do what you want.
 
Hi @FrankLen is this your first time here?
 
3:40 PM
@JonasTeuwen Why do you want to introduce references right now? You can take care of the references at the end after you have written everything. Just make a "\note" or a "\todo" in LaTeX wherenever you want to put a reference.
for "\todo"
\usepackage{todonotes}

\begin{document}
\listoftodos
.....
\end{document}
 
@JayeshBadwaik I know the package! I have my own fork.
 
@JonasTeuwen fork! Awesome! What did you change??
 
Because I already have those in my library, and if I had to do it at the end, I could just as well have included it.
@JayeshBadwaik Just more fine grained control of what I can make. I also have a list of remarks and stuff like this and answered/unanswered questions and so on.
 
@JonasTeuwen Hmm.
 
Mmm?
 
3:43 PM
@JonasTeuwen Okay! I have similar thing to, color coded lists with different colors and everything, but I created a new package for it.
 
@JayeshBadwaik I can include the key already.
@JayeshBadwaik Don't you think we need a versioning package that makes branches? 8-)).
If you remove the draft flag it should commit it.
 
@JonasTeuwen That can be done with commandline arguments and a suitable make file.
I already have that actually.
 
Oh??
So you can have like \advisorfriendly and then it will allow that in the branch advisor?
The rest in the main branch.
 
@JonasTeuwen No, not actually like that, I have a draft which does not commit and a main which commits.
 
That is pretty nasty with a makefile, it would have to parse the document.
How does it work?
 
3:46 PM
@JonasTeuwen Well make has two targets, the default is a draft and the special is a commit.
so when I type just make, the make sends the \draft command argument to pdflatex using the link I have shown.
 
How do you mark which one belongs where?
 
And when I type make final "message"
 
@JayeshBadwaik Seen this?
@JayeshBadwaik Oh, like that. But I have multiple versions also having crap in there which should not be public and only for me.
There should be multiple versions as in "Me", "Friends", "Whatever".
 
then you just define multiple make targets with appropriate commands in them
so for example
private: main.tex
git checkout private
git commit $1 #$1 is the message from the command line
public: main.tex
git checkout public
git commit $1 #$1 is the message from the command line
I had not seen the link you posted.
so reading it now.
 
Mm, yes, that works. Only, I have some zsh script that just does gch for checkout and stuff. Oh well.
Also the package gitinfo.
 
3:54 PM
@JonasTeuwen wow, that's nice. It is probably a good solution, because, if you want more features the commandline stuff will become really really ugly to handle. For one or two features, the commandline will probably save your time, doing the things the dirty way.
 
@JayeshBadwaik And this...
@JayeshBadwaik But I will use it for many years to come, so better do it right. Right?
 
Right! But now your paper........
My philosophy is take a dirty fix and use it for now, while you work on your script for later.
 
Yea... you're like my advisor 8-(.
 
@JonasTeuwen :P :P
 
@JayeshBadwaik I have asked.
 
4:03 PM
@JonasTeuwen Also, a big danger here is the overwhelming odds of a bad design (not software design but implementation design) on the first iteration, which ends up wasting a lot of time, when you are postponing other things (the things which the tool was supposed to speed up) due to this.
 
I know.
Then you delete and write it again.
I need something "higher" than \part 8-(.
 
@JonasTeuwen Whaaaaa?
 
Yes, as my main document includes my other documents which already have chapters.
I should just be allowed to give it a level, not like this.
Need more fine grained control!
So I have a part like introduction, but then I want the papers in their own part. Okay fine. But I want a SuperPart that is "Paper".
But perhaps, I should just push the levels of the other ones down.
As papers are okay with sections, and the main thing would be with chapters and parts.
 
@JonasTeuwen You are implicitly trying to write a content managing software bro (inside a publishing platform.)
 
Yea, kinda strange.
Anyway, I can just make a new one, no need CMS.
I already tried to do such a things... 8-).
It kind of works. My "documents" have no layouting!
 
4:09 PM
I gave up doing that two years ago, and have since shifted my attentions to quick fixes for now, as I slowly and painstainkingly design small bits for myself.
 
Just a collection of environments with names which are mapped to sections.
 
@JonasTeuwen Nice! I have been doing similar stuff, but its not very customized. I basically now just write small notes which I manage with mendeley. And for big papers, I just write the papers themselves.
 
Would be pretty awesome if I only have to checkout branch paper1 to get the paper that is included in my thesis.
Uh, pull it.
 
@JonasTeuwen hmm, yup, it would be awesome. Actually, that is one of the capabilities of knotfs. I mean not implicitly, but you can make it do that, by writing an additional tool based on it.
You can build a cms with just knots, basically.
 
@JayeshBadwaik I bet your thesis will end up being on in software engineering about a tool named knots.
With a nice section of applications.
 
4:14 PM
@JonasTeuwen :P :P No, I have something other in works (more general), but it is similar to that stuff, so probably yes. :P :P
I can have my dreams of producing "the ideal program". Can I not?
 
Of course.
Now go implement it. I need it.
I like small dedicated tools which do what the need do well good and what they don't need to do, they don't do.
 
@JonasTeuwen No pressure!! I have to pass my entrance exams first!!
 
What do you plan on working on?
I didn't do an entrance exams...
 
@JonasTeuwen Well, you were not in India.
 
Hmm... what is the relevance?
Did I have you in my facebook already? I believe so.
 
4:24 PM
@JonasTeuwen To get admitted to a graduate program, you have to pass an entrance exam.
@JonasTeuwen Yup.
 
Oh, here you need a job interview.
I skipped that as well.
 
@JonasTeuwen Yup, actually, if I had done a bachelor's in math and had a good enough grade, I could have continued to do masters at that institution without any of this.
 
What do you want to work on?
 
@JonasTeuwen Yup, I was going to respond to that before, but then, relevance question came in. I am not exactly sure which field since I don't have much knowledge as of now, but I am thinking either functional analysis (probably things similar to what you are doing) or might wander into theory of computation or something similar (quantum computation?). There are some fields like this I am interested in, depending on how the next two years goes, will decide on either one of them.
I have done some amateurish stuff in all the above fields, and some neat stuff in computation, so I am not decided yet.
One of the paths is something similar to this.
 
@JayeshBadwaik Nice article
@JayeshBadwaik I have done amateur stuff in all those fields!! 8-).
What kind of functional analysis?
 
4:38 PM
@JonasTeuwen My amateur level is probably currently your rookie level right now.
@JonasTeuwen Nice. :-)
@JonasTeuwen I have seen some stuff from Banach Spaces and Sobolev spaces in FEM methods. That was interesting. Spectral FEM and like.
Also, I have seen some quantum mechanics relations (but mainly wikipedia articles and papers applied to computational solutions, nothing real stuff), so I am curious about that. I am not really sure though.
 
@JayeshBadwaik I know quite a bit about that. If you want to chat about it, I am available for such conversations!
With the difference I don't care much about numerics, but I do care about approximation.
 
4:54 PM
@JonasTeuwen Great! I too also do not care much about numerics, but till recently I did not completely understand the concept of approximation or the difference from numerics, so did not appreciate it much. As for now, my aim is to be good enough in basic algebra and analysis and topology to pass entrance exams with good score. Till then, I might still look around, but probably not that vigorously.
 
You approximate your solution with things in a space where those solutions lie dense it basically.
So if it were to converge, it would converge to a "real" solution.
 
@JonasTeuwen Okay.
 
@JayeshBadwaik Also I use biber :-).
 
hi @Charlie
 
@infinitesimal hi!!!hi tiny friend!
 
5:05 PM
@everyone, any Aussie here? :)
 
@JayeshBadwaik Also, if you did not set up the "required" fields Mendeley makes a not-parsable .bib 8-(.
 
@JonasTeuwen Ohh. Never used it myself. Will try it sometime.
 
@JayeshBadwaik It removes some crap from bibtex and makes it more awesome.
 
@JonasTeuwen That's stupid.
@JonasTeuwen okay.
@Charlie Good afternoon.
 
@JayeshBadwaik Biber warning: [125] Utils.pm:168> WARN - I didn't find a database entry for 'NEEDS CITATION' (section 0) 8-).
 
5:07 PM
@JayeshBadwaik good night!!!how are you?
 
@Charlie i'm like the tiny bubbles in the wine that make you feel happy and make you feel fine...
 
@JonasTeuwen Haha. Nice.
@Charlie Hi! I am good! Good evening it is by the way. ;-)
 
@infinitesimal i do not drink...
 
Night is only uttered when you want to say good bye implicitly.
@JonasTeuwen If you add the following line to your sudoers file, it emits insults when you get your password wrong. Its great!

Defaults insults
 
@JayeshBadwaik what?vould i ask...have you been praying?
 
5:13 PM
@JonasTeuwen Jonas Jonas Jonas
@JonasTeuwen No, really. Are you there? This should be a piece of cake for you.
 
I am now!
 
@PeterTamaroff Now he will come, he heard "cake"!!
 
@JayeshBadwaik Oh! The best feature.
 
@JonasTeuwen Good. Riemann Lebesgue Lemma.
 
Additionally we need an insult daemon.
@PeterTamaroff Okay. Trivial. Done!
 
5:16 PM
@JonasTeuwen Yup. And a custom dictionary too!
 
I have proven that $f$ is integrable on $[a,b]$ $\iff$ for each $\epsilon>0$ there exist step functions with $s_1\leq f\leq s_2$ and $\int_a^b s_2-\int_a^bs_1<\epsilon$
@JonasTeuwen I have also showed the RL lemma holds for step functions.
So I wanna use the above to finish it off for any integrable $f$.
 
@PeterTamaroff Yes, then you are done.
 
@JonasTeuwen I can't get the argument together.
 
user19161
Sorry, internet connection problem...
 
@PeterTamaroff Oh. Shall I give it or do you need a hint?
 
5:20 PM
@Charlie you don't have to ... to listen to the song
 
user19161
@anon Are you feeling better now?
 
@JonasTeuwen Hints are always better =)
 
@JasperLoy yup
 
user19161
@PeterTamaroff Why do you need hints when the proof is in the book?
 
@PeterTamaroff Right. You want the result for Fourier series. Do you know something about their linearity and can you bound them by the function itself?
@JayeshBadwaik Available on FB?
 
5:23 PM
@JonasTeuwen Yes. Just a min.
@JonasTeuwen I am there now.
 
@JonasTeuwen Bound the step functions?
@JasperLoy It is not in the book! Spivak wants it proven.
@JasperLoy He made me prove the thing about step functions, and now this.
 
@JayeshBadwaik Ah!
 
Who starred that?
@JonasTeuwen Sooooooooooooooooooooo =) ?
 
@PeterTamaroff I gave a hint?
 
5:38 PM
" You want the result for Fourier series. Do you know something about their linearity and can you bound them by the function itself?" ?
 
Yep.
Like.
$|\hat s - \hat f| = |\widehat{s - f}| \leqslant \dots$
 
user19161
Just resumed my internet connection.
 
@JonasTeuwen $\hat s$ is?
 
Ugh. Such an ugly hat.
 
@PeterTamaroff Your task to figure out.
 
user19161
5:40 PM
@jayesh Now you are sharing secrets with Jonas on FB?
 
@JasperLoy Yup. The secret (sword) master!
 
@JonasTeuwen Oscillation?
 
Fourier series.
 
@JonasTeuwen But I'm not studying Fourier series...
 
Then what the heck is your Riemann-Lebesgue lemma??
 
5:45 PM
@JonasTeuwen Let $f$ be integrable on $[a,b]$. Then $$\mathop {\lim }\limits_{\lambda \to \infty } \int\limits_a^b {f\left( x \right)\sin \left( {\lambda x} \right)dx} = 0$$
 
Yes. So you are studying Fourier series of odd functions.
 
user19161
Now you know why I don't do exercises @peter.
 
@JonasTeuwen Well, suppose I were...
@JonasTeuwen The same result holds for $\cos$ right?
 
@PeterTamaroff The hatted version is this integral.
 
@JonasTeuwen Or the odd part of a fourier series...
 
5:49 PM
@robjohn 8-).
 
$$\eqalign{
& \int\limits_a^b {f\sin \left( {\lambda x} \right)dx} = \int\limits_a^b {\left( {f - {s_1}} \right)\sin \left( {\lambda x} \right)dx} + \int\limits_a^b {{s_1}\sin \left( {\lambda x} \right)dx} \cr
& \int\limits_a^b {f\sin \left( {\lambda x} \right)dx} = \int\limits_a^b {\left( {f - {s_1}} \right)\sin \left( {\lambda x} \right)dx} + \int\limits_a^b {{s_1}\sin \left( {\lambda x} \right)dx} \cr
& \left| {\int\limits_a^b {f\sin \left( {\lambda x} \right)dx} } \right| \leqslant \left| {\int\limits_a^b {\left( {f - {s_1}} \right)\sin \left( {\lambda x} \right)dx} } \right| +
@JonasTeuwen ?
 
00:00 - 18:0018:00 - 00:00

« first day (796 days earlier)      last day (4230 days later) »