« first day (2929 days earlier)      last day (2092 days later) » 
02:00 - 22:0022:00 - 00:00

2:22 AM
So I wrote a math question which is kind of eclectic: It draws from a few parts of math and I am not quite sure how to tag it. There are many appropriate tags for it and I am not sure which ones are the best to get the right people's eyes to the question.
 
link please
 
3
Q: A nicer form for $\sum_{r=1}^{m}\sum_{d|r}(-1)^{r+d} d^3$?

MasonSpecific Question Let $R$ be a whole number. Is it possible for any whole number $c>0$ that there is a nicer form for $\sum_{r=1}^{R^c}\sum_{d|r}(-1)^{r+d} d^3$? I can hope that we can just evaluate this sum but in lieu of that maybe something cheaper computationally. Taking $c=1$ we can exami...

Theta Functions doesn't seem not appropriate.
I just dropped "divisor sum" for the sake of Diophantine equations.
For the record: I think that this type of in-chat self promotion of your own question is not great form but probably ok if it only happens rarely. Hopefully this community agrees.
 
2:42 AM
i agree, it is certainly much better than the next option posted in the room, namely 4 hours later... :-)
 
 
1 hour later…
4:05 AM
hey is it true for holomorphic function we have the sum of the zeroes and the poles must be zero if we count multipicities ?
 
Holomorphic where?
Also I guess you mean meromorphic here. If you're talking about on any domain then no, consider $f(z) = z$ on the unit disk
 
Karim: That is very sloppily stated.
You mean the numbers of ... ? And on a compact Riemann surface?
and holomorphic map to $\Bbb P^1$ ?
 
4:27 AM
did you watch the Fields medals presentations Professor? @TedShifrin
Birkar's life is amazing
 
well I've seen something like that for modular forms @Adeek @TedShifrin
 
4:59 AM
Hello people! I must prove the following expression by induction: $\sin(x) + \sin(x+h) + ... + \sin(x+nh) = \frac{\sin(x + nh/2) \sin((n+1)h/2)}{sin(h/2)}$. I tried $\sin(a+b)$, some substitutions but to no avail; any suggestions?
where
n is a natural number and h
doesn't equal $2\pi n$
$\frac{\sin(x + nh/2) \sin((n+1)h/2)}{\sin(h/2)} + \sin(x+(k+1)h) = \frac{\sin(x + (k+1)h/2) \sin((k+2)h/2)}{\sin(h/2)}$
And I cannot equate the LHS to the RHS.
 
sorry @TedShifrin I was sloppy
on compact Riemann surface.
I am just reading something related to algebraic cycles
@user1732 @TedShifrin I am really happy for Birkar I had similar life originally living in that area of the world.
@Daminark @TedShifrin Riemann surface
I am trying to understand this statement $\Sigma n_i$ must be zero
 
5:52 AM
yeah the reason is like you said (applies to meromorphic functions on compact riemann surface)
 
I am struggling to grok the formula given in this oeis. Should it be a generating function?
2
what do these [] mean in expression [g(x)]? floor or ceil or something like this?
 
6:27 AM
@Adeek he has achieved so much starting from such a humble beginning.
 
institutionalized racism
 
practiced to this day
it makes Birkar's life even more amazing
 
7:25 AM
conjecture: let $f : (a,b) \to \Bbb R$ be differentiable and $f' > 0$ except for a finite number of points. then $f$ is strictly increasing.
 
that's true. Those finite number of points must be saddle points anyway
 
7:37 AM
A function is the integral of its derivative
If f' = 0 on a set of measure 0, and f' is nonnegative, then f is strictly increasing
 
7:56 AM
Cédric Villani gets on my nerves a bit
 
hi @loch
 
yeah me too, a bit
especially that spider pin he wears :P
 
8:22 AM
Hey everyone
Here's a question that I don't really have a clue how to begin
Find all solutions in integers of the equation $5x + 7y = 17$
 
8:40 AM
@Perturbative are you familiar with linear algebra?
 
Yeah @LeakyNun
 
then it is $\begin{bmatrix}5&7\end{bmatrix} \begin{bmatrix}x\\y\end{bmatrix} = [17]$
every solution is a particular solution plus a multiple of something in the kernel
 
8:59 AM
Hmm okay, not sure I fully understand what you're saying @Leaky, but there's a general solution using the extended euclidean division algorithm
Which I found through google
 
9:50 AM
@Perturbative you can first apply modulo $5$ both sides and conclude that $y = 5k +1, k \in Z$ and substituting back you will get $x = 2-7k$
Suppose $A$ and $B$ are matrices in row reduced echelon form and $A$ is row equivalent to $B$. Then $A=B$
Help me prove this
nevermind, got it.
 
10:07 AM
Hi @LeakyNun
 
Hello!!

We have the group $G=S_9$ and the elements $\sigma_1=(1 \ 2 \ 4 \ 5)$, $\sigma_2=(1 \ 2 \ 4 \ 6 \ 8 \ 3)$ and $\sigma_3=(1 \ 2 \ 5)(1 \ 2 \ 6)$ of $G$. I want to check which of the permutations are even.

We have the following:

$\sigma_1=(1 \ 2 \ 4 \ 5)=(5 \ 4)(5\ 2)(5\ 1)$ so this is an odd permutation

$\sigma_2=(1 \ 2 \ 4 \ 6 \ 8 \ 3) = (3 \ 8)(3 \ 6)(3 \ 4)(3 \ 2)(3 \ 1)$ so this is an odd permutation

$\sigma_3=(1 \ 2 \ 5)(1 \ 2 \ 6)=(1 \ 5)(2 \ 6)$ so this is an even permutation
 
Hey @LeakyNun !! Do you have an idea about my question?
 
ehh all correct
 
Great!! Thanks!! @LeakyNun
 
10:18 AM
hi
is anyone here familiar with polytopes?
 
10:47 AM
We have the group $G=S_9$ and the elements $a=(1 \ 2 \ 5 \ 9)$, $b=(4 \ 2)$ and $c=(3 \ 1 \ 4)$ of $G$.

If $d=abc$ then $d(1)$ is the following, or not?

$d(1)=abc(1)=(1 \ 2 \ 5 \ 9)(4 \ 2)(3 \ 1 \ 4)(1)=5$
 
@MaryStar yes
 
@LeakyNun Thank you!!
 
 
1 hour later…
11:57 AM
another link to Turkey :-/
 
Can anyone help me in simplifying $(n-k)\log [1-(n/k)] - n if $k \gt \gt n$
I am getting $n^2/2k$ however it seems that it should simplify to $-n^2/2k$
 
@user1732 watched the video
 
12:25 PM
Hi @all
Good Evening
 
12:43 PM
hi
long time no see?
 
I don't know you. Your name?
 
12:58 PM
skullpatrol
 
oh. hi
good to see you
 
good to see you too, pal
 
2
Q: Can these equations be considered as differential equations?

Rajesh DachirajuConsider a differential equation with a term containing $y(x_0)$, for example $$y'' - 2y' + y = y(x_0)$$ $x_0 \in \mathbb{R}$ is a constant. My question is, does such equations fall under the category of differential equations? I have never studied any equation with such a term. If its a differe...

 
1:18 PM
$$\left(\sin x+\cos x+1\right)^2 = 2\left(\sin x+1\right)\left(\cos x+1\right)$$
2
truly remarkable
 
1:41 PM
@user2236 There are many people like him in the world especially in the weird part of the world who are very gifted but didn't get the chance to develop their mind.
@user2236 I personally knew many in my home country. One person I know actually knew 5 languages by speaking with tourists. But because his ideas didn't conform to society norms his own family sent him to a mental institution lol.
2
 
@Adeek but over-coming such adverse conditions is rare.
 
Do you guys often find that people think that you talk about math all the time but you think that you don't? I get that all the time, and I don't know what to do about it.
 
Think about it more and talk about it less.
 
Yeah, but don't you ever have one of those days when you feel that math is so cool and you just can't stop thinking about it?
 
but I actually talk about math all the time
 
1:55 PM
I think that my problem is that I don't like that many people at the math department.
@Leaky What do you do for a living?
 
2:10 PM
i'm just a student in a university @OskarTegby
 
Okay.
 
@OskarTegby
what is your profile picture of? it looks really cool
@LeakyNun aren't you a student at one of the top universities in the world lo
 
It's just an octagon with a bunch of triangles drawn in it. It's a picture number four on this webpage.

http://www.thorbjornandersen.com/paintings-and-objects/geometric-objects/
By the way, have you guys heard about the YouTube channel 3blue1brown? It's so good. I love how he animates things.
 
@OskarTegby yeah its a great channel
 
Awesome channel
 
2:24 PM
I can't wait for them to make one on metric spaces
 
Yeah! I really love how he actually teaches how to approach math instead of just looking at one cool mathematical object.
Are you guys also at uni? What do you study?
 
vzn
what is annoying about villani? too flamboyant maybe? havevnt watched his videos yet. liked this book, recommend it, very accessible. birth of a thm amazon.com/Birth-Theorem-Mathematical-Cédric-Villani/dp/… then theres this too. MEET THE ‘LADY GAGA OF MATHEMATICS’ HELMING FRANCE’S AI TASK FORCE theverge.com/2018/3/28/17170104/…
 
@OskarTegby i just graduated with a degree in stats but I'm planning on going back for a master's in mathematics in a year or tow
you?
 
I'm starting my masters in mathematics this semester.
 
nice
what classes are you takin
 
2:40 PM
Integration theory, functional analysis, and analytic functions.
 
good luck.
 
Thanks!
There are so many courses that I'd want to take, but I don't have any time.
If I don't do a PhD I'll probably try to keep taking classes after I've graduated. What sane person would say no to learning more math?
 
yeah i understand. I know that Richard Feynman took like every physics class at MIT when he was an undergrad lol
yeah i know math is awesome and there's always so much more to learn!
 
Yeah, but learning doesn't need to be limited to classes or even university. After having been to university one's already learning so independently that there are no limits.
I would like honestly be fine with selling hotdogs and doing math.
 
haha that's when you know you have a passion for math
if you don't need tons of money to want to do it
 
2:55 PM
Yeah!
 
0
Q: Finding a metric in $ \Bbb R^2 $ depending on $s$ such that $x^s+y^s=1$ is a geodesic wrt. the metric

George ThomasLooking for a complete regular Riemannian metric in $ \Bbb R^2 $ depending on $s$ such that $x^s+y^s=1$ is a geodesic wrt. the metric, $x,y\in(0,1), s\in \Bbb R(1, \infty). $

@OskarTegby can you explain this to me?
 
Maybe. What is it that you don't understand?
 
Why do you need a change of coordinates if the curve already goes through the first quadrant of the unit square
 
3:10 PM
Does it, though?
 
3:23 PM
Hi people!
 
@Rudi_Birnbaum hi
 
Hi @Akash.B!
 
@Rudi_Birnbaum Can we draw a straight line on earth's surface?
earth is spherical ,right?
 
yes, so probably no?
 
yeah I think so
@Rudi_Birnbaum then on what basis Euclidean postulates where formulated?
 
3:34 PM
Earth is actually flat
 
@Akash.B pardon? I don't understand your question
 
@Lozansky huh?
 
is it? must be some new development ...
 
@Rudi_Birnbaum oh?
 
Anyone here with a Erdős number?
 
3:37 PM
@Rudi_Birnbaum what sort of number is that?
 
The Erdős number (Hungarian: [ˈɛrdøːʃ]) describes the "collaborative distance" between mathematician Paul Erdős and another person, as measured by authorship of mathematical papers. The same principle has been applied in other fields where a particular individual has collaborated with a large and broad number of peers. The American Mathematical Society provides a free online tool to determine the Erdős number of every mathematical author listed in the Mathematical Reviews catalogue. == Overview == Paul Erdős (1913–1996) was an influential Hungarian mathematician who in the latter part of his life...
 
4:07 PM
I think most people who publish with a coauthor have one.
 
4:29 PM
@OskarTegby yes it does
every curve goes through the unit square
 
Unit circle. Right?
 
well the equation is $ x^s+y^s=1$
so for each $s$ each equation goes through the unit square
 
Okay. Yeah, you meant like that.
 
its an equation of super ellipses
a family so to speak
 
Yeah.
Have you tried to follow the proposed method? Maybe that will give you some insight.
 
4:34 PM
Yeah I'm currently working on it
thanks
 
Okay. Tell what you learn. I'll think about it.
I think that you're right, but the remaining part of the answer is probably of more interest.
 
okay sure
yeah i'm looking at the metric part
@OskarTegby do you know what a regularity requirement
is?
 
4:53 PM
From Wikipedia: "A differentiable curve is said to be regular if its derivative never vanishes."

Source: https://en.wikipedia.org/wiki/Curve#Differential_geometry
 
hm okay
 
As in $f'(x)=0$
That doesn't happen if it's regular.
 
oh no..
oops
 
If you look at the comment by Robert Bryant on the question you see why they asked this.
Also look at George Thomas' comment on the answer!
 
@geocalc33 The demand here is that the derivative exists everywhere and is continuous. This is not true for the curves you are interested in, so it won't be true of any metric that has them as geodesics.
 
5:03 PM
...?
 
That is what 'regularity' means in the comments on your question.
 
I thought the derivative does exist and is continuous in the open unit square
...
wrt to s
 
I'm tryin to solve the following problem.

"The integral $\int_0^\infty\sin x^2dx$ is called a \textbf{Fresnel integral} and it arises in wave optics. Show that this integral converges, by proving that the sequence $a_n:=\int_0^n\sin x^2dx$ converges in $\mathbb{R}$."

It's from an exam in real analysis so it shouldn't require tons of ugly computations, but all I find in the below thread involves so many weird things. Is my intuition wrong? Is there an easier way to do this in my case?

Thread: https://math.stackexchange.com/questions/187729/evaluating-int-0-infty-sin-x2-dx-with-real-methods
It comes with the following hint.

"Use the fact that
$$
\sin x^2=-\frac{1}{2x}\frac{d}{dx}\cos x^2.
$$"

I've tried integration by parts, but it doesn't get better.
 
@MikeMiller it does vanish... sorry
it vanishes inside the unit square
but the derivatives still exist outside the space
grrrrr
 
The problem is sign trouble at the axis
The equation you really mean is $|x|^s + |y|^s = 1$ (hard to know what $(-1/2)^{1/4}$ means)
 
5:14 PM
oh okay
maybe ill change that
 
But that makes the derivative issues at the axis points $(\pm 1, 0)$ and $(0, \pm 1)$ even more clear: think about the case $s = 1$
You're just a diamond then, and the derivative has a jump at the axis points
 
I wish i could find a way to just ignore those points...
 
The answerer did this by restricting just to the first quadrant
As long as the curve stops at the axes, you don't have any claim that there's a problem with jumps in the derivative
So you think of the first quadrant, and it is nice and easy to parameterize this in terms of the variable $[0, \pi/2]$ (angle) and $(0, \infty)$ (radius)
They first found a metric in which the unit circle was geodesic
And then found a map that takes the unit circle to your circle
One can pull back a metric under diffeomorphisms, and geodesics of the pull-back metric are the inverse image of geodesics in the old metric. so doing so gives you the desired metric (in the first quadrant)
In general you could glue that to flipped copies of itself for the other quadrants but derivatives won't line up at the axes so it'll be a continuous metric but not $C^1$
 
so as long as the curve stops there's no problem with jumps in the derivative
okay i think i figured it out
thanks
 
https://math.stackexchange.com/questions/2879581/show-that-int-0n-sin-x2dx-converges

Go get that reputation!
 
6:00 PM
Example of some insane topology:
Let fascist free ideologies be open sets. Then fascism is closed by complement
Then anarcho-capitalism is an open set whoose limit points is located within fascism
 
I can't tell if serious or not
 
I spent a whopping 13 hours analysing this thing using some fascist checklist
I am dead serious because I like anarcho capitalism, but my left wing friends will not allow me to post it if it contains fascist ideas
Thus I spent 13 hours just to prove that anarcho-capitalism is fascist free, and then nature smack me with this
$$\overline{\text{anarcho-capitalism}} \cap \text{fascism} \neq \varnothing$$
that's how deadly close ancap is to fascism
(PS I spent 10 out of 13 hours in order to locate where Trump is as well his future evolution)
Anyway, the actual limit process is: The ideology is fascist free, but nearly all its thinkers and supporters are not
So basically, unless you stick to wikipedia, or maybe the ancap reddit, you will be screwed
 
6:22 PM
@OskarTegby: There seem to be answers to your question. If you have a specific follow-up question, let us know.
 
@LeakyNun Needed help in spotting error in solution. Are you there?
never mind...
 
6:48 PM
Hi @Ted
Hi @Paul long time no see!
 
@Alessandro Hello
It has been a bit
 
Hey @AlessandroCodenotti
How's the alg top going?
 
What kind of math have you been dealing with? @Paul
@Perturbative Uhm I got sidetracked by the de Rham cohomology
Which is still AT I guess, but I was supposed to learn singular homology first
 
How are you finding it learning de Rham first?
 
2
Q: $\dfrac{\varphi(6)}{6} \cdot \dfrac{\varphi(5)}{5} (7^2 - 5 + 1) =$ number of primes $5 \lt x \leq 7^2$.

EnjoysMathA sufficiently large prime number $n$ can only be in certain residue classes modulo another integer $m$. For instance modulo $6$, these classes are $1, 5$. The number of classes precisely is $\varphi(n)$ or the value of Euler's totient function. Can we use this information in prime number prob...

 
7:02 PM
@AlessandroCodenotti geometric group theory. Been looking at some stuff on the Flat torus theorems, which basically says that if you group acts on a space satisfying nonpositive curvagture conditions, then abelian subgroups have to come from euclidean flats in the space
 
I did learn a bit of homology first, but not as in depth as I should have. I'd suggest going the usual route, homology first, cohomology later, but I don't know nearly enough to give pedagogical advices in AT!
 
How about you @AlessandroCodenotti
 
I finished my bachelor in July and I'm catching up on some algebraic topology now because we have very little AT in our undergrad syllabus while the uni I'm going at for my Master has two undergrad courses in AT
@PaulPlummer Not my cup of tea but it sounds cool
 
What is your cup of tea? Congratulations on finishing bachelor
 
Set theory, logic and related stuff, especially model theory
 
7:09 PM
Set theory is great, I think if I did go down the set theory hole it would probably be more of a descriptive set theory route though
 
I started reading the first chapter of Kechris for an undergrad talk project we had this semester at our uni, but then I changed topic and talked about cardinalities of ultraproducts instead
The little descriptive set theory I've seen was super interesting though, I definitely want to learn more about it
 
How can such a large empire have only ~20 common first names
 
@AlessandroCodenotti Yeah I'm gonna learn singular homology in a couple months time in my alg top course, hoping to learn de Rham after that
How was the process for applying to study in another country btw? Was it as stressful as it sounds like
 
Nah, they're pretty cool in Bonn
They accept more or less anything as a proof of proficiency in English for example, doesn't need to be the result of a standardized test (TOEFL or stuff like that)
 
That's nice to hear
 
7:24 PM
I just had to send them a transcript of my courses, a statement of purpose and a letter of recommendation written by one of my professors
 
Damn makes me wanna apply to Bonn when I finish my bachelors
 
Being an EU citizen makes the bureaucracy way easier though
 
Silly question but are you gonna take undergrad courses (like the alg top course you mentioned) while doing your Masters?
 
Not if I can avoid it
 
Ahh okay cool
 
7:31 PM
You can only do one or two undergrad courses and receive credits for it, but they could be in German instead of English and it delays the grad courses in the same topics so it's a bit annoying
 
@AndrésE.Caicedo Hello!
 
I'll see how that turns out, but I hope I won't need to take undergrad courses
@AndrésE.Caicedo Hi!
 
At the very least even if you do it'll be in something like AT instead of numerical analysis
Which for me would be a good trade-off :p
 
I completely agree!
 
8:02 PM
Hey @Dami
 
Hey @Alessandro and everyone! How's it going?
 
I accepted on fb
 
I did notice :P
Well, now that I've sent my current paper draft to my mentor, I've gotta find something else to procrastinate on GRE studying with
 
8:22 PM
Why do you need to take the GRE?
 
Whats the paper on?
@Daminark Is Danny your mentor?
 
@PaulPlummer Hi Paul. How are you?
 
@AndrésE.Caicedo Doing good, cool to see your around. I just arrived in Toronto a couple days ago and will be spending a semester at Fields! How are things with you?
 
@PaulPlummer no, my mentor is one of Emerton's grad students. The paper is on dinner interactions between Galois theory and elliptic curves
@Alessandro to go to grad school in America you have to do the GRE :/
 
Ah, I wasn't aware
Come to grad school in Europe, we're nice
 
8:30 PM
Ah okay @Daminark
 
can I come too
America sucks
 
Lmao, I could consider it, though I've been cautioned that you have less time to do research in Europe which can make life a bit more difficult when it's time to apply to postdocs. Also funding is trickier as an international
Though on the other hand healthcare so I dunno
And no GRE :)
@Paul are you still doing GGT?
 
@Daminark Studying for the general test or the subject test?
 
@Daminark Does Canada require the GRE?
 
im sure you'll be fine with your GRE --- make sure you do some practice tests!
 
8:33 PM
@Fargle Sure, there's room for everyone
 
Hmm, I don't think they do, I could consider that.
And subject test, just gotta learn how to compute integrals again
 
@Daminark Yah, mostly
 
I'm mildly relieved that it seems like in two of the practice tests, only 6 questions were on multi, so I guess that's something
@loch thanks for the advice and encouragement! I'll make sure to do so
 
@PaulPlummer Ah, that's excellent. Hope you enjoy it. Lots of good set theory in the area, too. I am doing well, enjoying Ann Arbor. (Also, I tore some tendons in my right knee in an accident a while ago. Had to have surgery, it was a bit traumatic. I'm still going to physical therapy.)
 
8:58 PM
@AndrésE.Caicedo I am glad you are enjoying Ann Arbor. I hope the injury recovers, knees are too delicate for all the things they do... Also, thanks for being such a great teacher!
@Daminark Still considering GGT?
Hi @MikeMiller
 
9:37 PM
@PaulPlummer definitely would like to try it out at some point in the near-ish future
 
@PaulPlummer Thank you, Paul. That's very kind.
 
@geocalc33 How did it go?
 
making progress! :)
I've gotten 200 views and nobody has said anything lol
 
The story of my life.
Why did you post on MathOverflow?
Maybe that's why.
 
why?
 
9:52 PM
Isn't MathOverflow less active than MSE?
 
Yeah but i had to choose one
maybe it'll be moved over to SE
 
02:00 - 22:0022:00 - 00:00

« first day (2929 days earlier)      last day (2092 days later) »