« first day (2283 days earlier)      last day (3033 days later) » 

00:03
[Complex analysis] apparently there can be no analytic functions with periods 1 and i because bounded analytic functions are constant?
@Shimmy The normal letter, $l$, looks too much like a $1$, especially in handwriting
hey @AkivaWeinberger
Hi
@Shimmy The real question is why is there a $\wp$
00:53
@Hailbert Use for what? You can't "use" literally most of mathematics.
It'd just be a bit more boring if every top. manifold was triangulable.
The more stuff in TOP starts to fail, the more the fun!
@DHMO Yes.
is there a template for arXiv papers?
Because it'd be bounded in the square spanned by 1 and i, by compactness of the square, and then periodically extended that'd also be bounded.
they all have like green and red outlines for references and this specific abstract size
01:19
@TedShifrin I found an elegant (no sequences) proof of the oscillation thing.
 
2 hours later…
02:50
Still no progress here
1
Q: Solving for best fit value $C$ in $\sqrt {Exp_a^{[1/2]} (x) \cdot Exp_b^{[1/2]} (x )}$ ~~ $ Exp_C^{[1/2]} (x).$

mickLet $Exp_t^{[y]} (x) $ denote the $y$ th iteration of the exponential function with base $t$ : $t^x.$ For example $Exp_t^{[1]} (x) = t^x. $ Let ~~ denote best fit. Now as $x$ Goes to positive infinity and a pair $(a,b)$ with $e<a<b$ Is given , I wonder how to find the best fit base value $C$ ...

Yes that is a tetration question
The horror :)
Hello
 
3 hours later…
05:52
I need a favor
Could someone define mathematics for me?
06:39
@SirCumference That's a hard job to do. Why do you need a definition?
07:10
@BalarkaSen It'd make a killer introduction for a speech I plan to give
 
2 hours later…
09:05
"Mathematics is rigorous reasoning about abstractions."
But, that still doesn't capture a whole lot of things I still consider "mathematics".
09:26
Define cooking, and ask a cook/chef if that is ok for him/her. (prolly not exactly)
Hi @Fargle
for a 4 year old mathematics might be counting to 10(or 100), and that would ineed be something to be proud of. Nontheless is counting not all, neither is proofing or implementing.
Howdy @Balarka
Just woke up, as a matter of fact. Yourself?
09:29
not much, been busy with a lot of nonmathematical things today
Sounds like a drag.
tell me about it...
thats a hard thing to do, where we know from the show "Numb3rs" that all is math. (running)
i feel like that show is a parody.
If I had to explain to someone what mathematics is/mathematicians do, I'd say it's about studying numbers and pictures.
@saturatedexpo My favorite part is when they "explain" IRC.
09:31
But that doesn't do a good job without more context.
Eg I'd probably explain that statement further by giving examples if I was talking to them.
math is not (all) about numbers (insert here what you like), like sex is not all about orgasms.
The definition I usually give laymen (it's funny because I'm one too, strictly speaking) is "the science of patterns and change".
i could live with that (as a fellow laymen)
"Science" isn't quite the word, but it does a good job.
maybe "art"
"art of recognizing patterns and change"
09:40
I don't know if "art" is quite right either. Sure, there is a palpable aesthetic nature to it, but that isn't the primary function of math, at least to me.
@saturate I didn't just say numbers. Generally the kind of math I like is described moderately well as "study of pictures". But of course it's hard to explain, let alone define all of mathematics in one line.
mmh, i mean more like in "art of war", something to master, but nothing much aesthetic
A perhaps deeper definition might be "the search for formal truths about patterns and change".
I see. I can agree with that.
altho some proofs i seen where art.
Heck if someone can explain a concept in "downtoearth"-language, i'd consider it even aesthetic art
@saturatedexpo There's no question that beauty is a great motivator for mathematics--beauty in elegance, beauty in simplicity, and beauty in depth.
09:44
i think showing someone the mandelbrotset(some of those cool images), and saying math contains this, would be ok.
For me, mathematics is the study of formal structures, and their behaviour
Even though I don't quite understand the nature of the connection, monstrous moonshine has always struck me as having the beauty-in-depth quality.
@Secret I can jive with that, too. I like that mathematics is one of those things that has many (I guess) satisfactory, non-equivalent definitions.
this is because mathematical objects can be the subject of inquiry, a toolbox to investigate the subject of inquiry, and "atoms" and "laws" that govern an abstract world
i guess that would also prepare them in the best way to begin studiying math, to show them that the mathematical approach isn't always the same.
like sometimes drawing a triangel(altho a cleaver one) gives you fame over centuries^^
@saturatedexpo There are also examples (which escape me atm, very tired) of mathematics that is nothing like what most people think of mathematics.
And I don't mean, like, category theory.
09:48
Music @Fargle
I mean lines of reasoning that seem so common sense it feels strange to call them mathematical.
Or, indeed, music.
maybe a good counterquestion would be to ask, what they think is "music"
because music is not just all jazz,rap,pop,classic,..., neither it is +all instruments,...
@saturatedexpo To me, music is just sound with a context.
music is a combination of soundwaves given a subject meaning, which is why it is one of the few things that lies at the intersection of art, linguistic and science
to me(advocati diaboli) music is lady Gaga
lets see what wikipedia has for us
lots of words which probably have circular definitions. (like someone might have guessed)
09:57
@saturatedexpo Every word has a circular definition, I'd say. That or an infinite regression of definitions.
then there's the problem we where stitched from. Of course you can't define math in words satisfieing everyone.
i would even go as far that we can't define anything(placeholder) by using only a finite set of mediums(words, pictures, music...)
@saturatedexpo Which is exactly why many treatments of set theory leave "set" as an undefined primitive.
that's a little bit deep for 11 am
It's far too deep for 5 AM as well.
would have expected something like that for 11 pm haha
10:07
This IS the Philosophy of Mathematics chatroom right? ;)
@Fargle Even easier, the notion of a "point" in Euclidean geometry.
@BalarkaSen Quite right.
And "line".
Line is easy in projective geometry. It is defined by 3 collinear points
in euclidian its not that clear what you mean by coliniear tho
10:09
@Fargle I think you can define line as a pair of points without hurting anybody.
@BalarkaSen I suppose.
If line is a pair of points, then $S^1$ will become a line
because $S^1$ is literally a pair of points
that's nonsense
a point if something i get doing my homeworks
a line is something a snort
*i snort to be able to cope with this stuff
what have i missed, isn't the 0-sphere just two points?
10:12
S^1 is not the 0-sphere
ok sorry, should be $S^0$
Cannot count well because of the terminology of spheres and balls vs spheres in daily language
my gf studies psychology
in her book it says on the first page that there exist dualism and materialism. Hint: materialism is not what it is in common usage
reminded me a lot of "do numbers exists"
@BalarkaSen what I mean is that if people will prove new theorems, they most likely won't use that particular counterexample in their proofs
@Fargle The obvious trouble is that you then have to define when a point belongs to a line. Equivalently, define when you'd consider two lines to be "the same".
if instead every manifold was triangulizable, it would be used A LOT
10:22
ducks
if you know what i am saying
I think all of this stuff (Euclidean geometry robbed of the idea of lengths) is studied in incidence geometry.
Can you recommend a book to understand the connection of submanifolds, intersections and the cup product?
@Hailbert The paper does not just "give an example". It has a lot of new ideas, which could potentially give more ideas.
10:25
of course
i was referring solely to the theorem
not the proof
What's everyone's favorite proof that $\sqrt{2} \notin \Bbb Q$?
I rather like the "let $m \in \Bbb N$ be minimal so that $m\sqrt{2}$ is an integer, and consider $m(\sqrt{2}-1)(\sqrt{2})$" argument.
@Hailbert Even then, I do not see how you're so confidently saying people "most likely wouldn't use that particular counterexample".
@Fargle I like the proof via irreducibility of $x^2 - 2$ over $\Bbb Q$.
@BalarkaSen Oooooh, that one is rather clean.
How do you prove the irreducibility of that polynomial without already knowing that $\sqrt2$ is irrational?
Gauss lemma.
is it alright to disclose questions about MAT (mathematics admission test) here?
11:12
@BalarkaSen, doesn't any proof of that have to include something an awful lot like the "ordinary" proof of the irrationality of $\sqrt2$?
So, at least one of $\pi+e$ and $\pi-e$ is transcendental; at least one of $\pi+e$ and $\pi e$ is transcendental; at least one of $\pi-e$ and $-\pi e$ is transcendental.
Is it true that $a$ is transcendental iff $-a$ is transcendental? I think so because you can always substitute $x$ with $-x$ in the generating polynomial...
Yes, because it is true that $a$ is algebraic iff $-a$ is algebraic.
So let's construct a table of transcendentalness.
pi+e pi-e pi*e
no   yes  yes
yes  no   yes
yes  yes  no
yes  yes  yes
so at least two of $\pi+e$, $\pi-e$, and $\pi e$ are transcendental
@BalarkaSen I never can get my head around why bounded analytic functions must be constant... like, it isn't the case in $\Bbb R$
@MeesdeVries Something like that. But it'd not be a proof by contradiction argument.
@DHMO Yes, its false in $\Bbb R$.
Complex analytic functions have a bit more rigidity to them than real analytic ones. Not every real analytic function on $\Bbb R$ extends to a complex analytic one, precisely for the reason you said.
Ignore the last statement.
11:41
Well $R\in C$, so every function in R has to be also in C or not?
$R\subset C$ i mean
@saturatedexpo Every function $\Bbb R \to \Bbb R$ extends to a function $\Bbb C \to \Bbb C$. But not every analytic functions extends to an analytic function.
@BalarkaSen could we have an example?
$1/(1+x^2)$.
Hi @AlexClark
user147690
Hey @BalarkaSen
user147690
I'm just trying to pick my honours thesis topic for next year atm.
11:48
@BalarkaSen could we have an intuitive explanation on why analytic bounded functions must be constant?
user147690
What are you working on @BalarkaSen?
@DHMO I don't know off the top of my head, to be honest. The point is you have the Cauchy estimates there.
@BalarkaSen I dont think the first statement is true
@AlexClark mostly differential geometry. a bit of topology.
@DHMO Of course it is.
why?
11:55
You can explicitly construct an extension.
oh, alright
Just define $\Bbb C \to \Bbb R \to \Bbb R \to \Bbb C$, where the first map is projection, the last map is inclusion.
is there any power series convergent to absolute function?
namely?
12:00
You should think about it.
i'm thinking about its derivative which is a simple step function
that strikes me as similar to arctangent
Good ideas.
so I would normalize it and raise it to the power of a big odd number
wait, bad idea
Ok, so you're approximating the sign function by $x^{1/n}$ where $n$ is some big odd number. That's fine.
I should raise it to the reciprocal of a big odd instead
12:03
Still need to work on a power series though.
oh, so arctangent isnt needed?
x^(1/n) is just exp((ln x)/n)
right approach?
Not sure what you'd do with that.
$x^{1/3}$ is eg not differentiable at $x = 0$ so you can't actually Taylor expand.
hmm...
12:07
Back to the drawing board? :)
You had good ideas (these are the ones I initially had), but there's an easier one to get away with for $|x|$.
Think square root.
...
sure, it is sqrt(x^2) ...
and then?
Smooth it out near 0. Pick some very small $\epsilon$, and look at $\sqrt{x^2 + \epsilon}$.
These approximate $|x|$, and are analytic.
hmm...
sqrt(x^2+e) = (1/n)sqrt(1+(nx)^2) ?
12:14
I don't understand the question
oh, n=1/e
sure, why not
so e->0 means n->infty
and then binomial theorem?
yep. or equivalently Taylor's theorem
sure
thanks
12:17
No worries. In general, a surprising theorem in analysis says every continuous function on a closed interval can be approximated, uniformly, by polynomials.
only continuous?
yeah!
This is the Stone-Weierstrass theorem.
but... what about the derivative of absolute?
You can do it for those too. Tweak your earlier x^1/n examples a bit.
Anyone knows a reference for the naturality of the poincaré duality isomorphism with respect to orientation-preserving maps?
12:19
but it isn't continuous
@DHMO Ok, good point. The tweaked approximation I have in mind is not uniform, only pointwise.
Sorry about that.
what if i just take the derivative of the power series earlier?
Nothing would give you a uniform approximation.
@abenthy That's just naturality of the cap product.
what means uniform?
See e.g., Hatcher chapter 3.3 for that.
@DHMO lmgtfy
12:23
alright
@BalarkaSen Right, thank you
The orientation-preserving bit is needed to preserve the fundamental class, is the point.
OK, I gotta run.
what i have read was: there exists a Bijection between $\mathbb{N}^{\mathbb{N}}$ and $\mathbb{R}$. But how to even imagine $\mathbb{N}^{\mathbb{N}}$
Is $\mathbb{N}^{\mathbb{N}}$ just $\times_{i\in \mathbb{N}}\mathbb{N}$?
12:39
loose and handwavey notations
I am trying to write the cartesian product of $\mathbb{N}$, $\mathbb{N}$ times
I'm not criticizing your notation
Either way, I have no idea how to type the big X
what is the big X?
Oh.... I never saw the big X so I don't even know it exists
I have always seen small X because I am reading (learning) everything from the internet
20
Q: How can I get a big cross to denote a generalized cartesian product?

Fernando MartinI'm trying to get a big cross which I can subscript in order to denote a generalized cartesian product (much like how \bigcup works for generalized unions). How can I accomplish this?

Apparently it is not possible
@Secret $\displaystyle\prod_{i\in\Bbb N}\Bbb N$
12:50
Ah I see
 
1 hour later…
14:17
my tutorers used $(M^{},\circ )$ if they wanted to notate the set M without those Elements that would "destroy" the Group. Like 0 in multiplication. So $(\mathbb{R}^{},\cdot )$ would be simply $(\mathbb{R}\backslash {0},\cdot )$. But one could find bigger exclusions easily. My question is: is there a more international (yet short) notation for "exclude all elements that would destroy this group"
use \ast for asterisk
chatroom will use the * symbol for italics which screws up latex
ah ok
yeah, not so much in chatjax again^^ 60 seconds is like nothing
@saturatedexpo you can't destroy a group you haven't defined yet, so maybe that's not the best way to phrase it. if M is a monoid, then $M^\times$ might be used to denote the group of invertible elements. this applies in particular with the multiplication operation in any (unital associative) ring.
some people use $M^\ast$ instead of $M^\times$
ah my used indeed $M^{\times}$ (forgot that little detail)
but they said, that can mean different things in different contexts
$M$ can be different types of things in different contexts, but I can't off the top of my head think of another usage for $M^\times$. I've even seen $V^\times=V\setminus 0$ for vector spaces $V$, which is a slight generalization.
14:31
so i can't per se say that: $(M^{\times},\circ )$ is a group by definition. Since the cross is only ambiguisly defined.
lets define $M^{\times}$ as the subset of M for which following holds: $(M^{\times},\circ )$ is a group. would that be ok?
where $\circ$ is a placeholder (it could be addition for example)
what's M?
a generic set, from emptyset or bananas and horses
then what's $\circ$? not something defined on $M$?
14:39
(R,+) means the reals under addition or not?
yes it means that
perhaps you want $(M,\circ)$ to be a monoid, then $M^\times$ can be the largest subset for which $(M^\times,\circ)$ is a group sharing the same identity element as $M$.
note the last condition is necessary. for instance there are subsets of $M_n(\Bbb R)$ (that's $n\times n$ real matrices) which are groups under multiplication but do not have the identity matrix, e.g. $\{{\rm diag}(1,0,\cdots,0)\}$
to be honest i think i just wasted both of our time
no i got it right, ok.
Am I right to say that a homeomorphism $f \colon M \to N$ of orientable topological manifolds with orientations $(\mu_x)_{x \in M}$ and $(\nu_y)_{y \in N}$ is orientation-preserving iff $f$ maps $\mu_x$ to $\nu_{f(x)}$ in homology?
also is there a relation to the elements $\mu_K \in H_n(M, M \setminus K)$ which restrict to $\mu_x$ for all $x \in K$ for compact sets $K \subseteq M$?
I mean, if $f$ is orientation-preserving, does $f$ map $\mu_K$ to $\nu_{f(K)}$ for all compact subsets $K$ of $M$?
@arctictern $\circ$ would be a placeholder for a generic operator, as example addition or multiplication.
maybe theres a better symbol for that
in order for this sentence to make sense,
9 mins ago, by saturatedexpo
lets define $M^{\times}$ as the subset of M for which following holds: $(M^{\times},\circ )$ is a group. would that be ok?
you need $\circ$ to be defined on $M$ first
14:45
Ah, otherwise you define something with undefined stuff, makes sense ;)
So a rigorous definition of a generic operation is rather hard for 1 hour in class?
if you don't specify that $\circ$ is unital and associative to begin with, then crazy things can happen
for instance the loop of unit octonions has many subgroups isomorphic unit quaternions, all of which are "maximal" in the sense they can't be extended to bigger subgroups, but there's no reason to prefer one subgroup over another so there is no uniquely defined subgroup
and that has an identity element
but im also right that $(R,\cdot)$ is no group, because 0 has no inverse?
correct
(where R is a ring)
oh, i mean $\mathbb{R}$
applies to any ring
14:50
i didnt know reals where "only" a ring, i thought they where a field.
I didn't say they were only a ring
ah ok, i just read it that way
$\Bbb R$ is a linearly ordered set, metric space, field, etc.
the only property that's relevant to the conclusion is that it's a ring though
cuz it applies to any ring
so there are some rings that are no fields and have still groups. and there are some fields, which dont have those same groups.
what do you mean by "some fields which don't have those same groups"?
14:53
ah, hard to say
if G is the group of units of a field, then any finite subgroup is cyclic. this is not always true if G is a group of units of a ring.
that's a thing you can say
I've read some pages (some of those that I can read) from Chen Nanxian, Mobius Inversion in Physics, World Scientific 2010, as a Google Book, and I recommend you, section 1.3.4.2 in pages 22 and 23 because seems to me incredible. Enjoy with it, @user1952009
@arctictern i guess what i wanted to say is: a field is not always better then a ring. But now you still can only guess what i mean with better. Got to go, but that will rustle my jimmies haha
Let $(V,\|\cdot\|)$ be a normed vector space. Is it true that, if $\{x_1,\dots,x_n\}$ is a linearly independent subset of $E$ then $\|x_i\|\leq \|x_1+\dots+x_n\|, \forall i=1,\dots,n$?
no
consider $V=\Bbb R^2$ and $n=2$. pick two vectors at an almost 180 angle of the same size
it is true if the vectors are orthogonal though
15:02
@arctictern not sure if $E$ means what you think it means
didn't even see the letter E
dunno what it means
I think it means (0,0,1,0,0,0) etc
the coordinate basis? hmm.
Oh, sorry!
E=V
Typo
Anyway your example did help, @arctictern. Thank you!
15:17
Am I correct that the taylor series of exp(x) is pointwise convergent but not uniformly convergent?
Hello could someone help me solve (z+2)^n - (z-2)^n = 0 for z where n is natural and z complex
@user379685 (I'm only trying to help but I don't know the solution) does it help to set it as |z+2| = |z-2| ?
Given a commutative ring $R$ and an $R$-module $M$, can you call another $N$ module free with respect to $M$ if it is isomorphic to a direct sum of $M$'s and get a reasonable notion?
15:32
0
Q: Congruence involving prime numbers

sashaGiven the function $f(x)= x^{\frac{p(p-1)}{2}}-1.$ Also let $p$ be an odd prime number. If $\epsilon$ is a number $\pmod{p}$ such that $f(\epsilon) \equiv 0 \pmod{p}$. Then how do I prove ( prove because Hardy and Wright use it but don't prove it ) that $$ f(\epsilon)\equiv 0 \pmod{p^2}$$

any help is appreciated
thanks
Is it frowned upon to make two edits within a short period of time?
I need to correct a grammatical error I just found...
let z=a+bi.
(a+2)^2=(a-2)^2
a=0
the equation becomes (2+bi)^n - (bi-2)^n = 0
((2+bi)(bi-2))^n - ((bi-2)(bi-2))^n = 0
(-b^2-4)^n - (bi-2)^(2n) = 0
now 2+bi and bi-2 has the same radius.
Let 2+bi=re^(iu). bi-2=re^(ipi/2-iu).
niu = nipi/2-niu+2ikpi
u = pi/4 + kpi/n
don't ask me how to continue; i have no idea
@teadawg1337 no. if you make two edits within 5 minutes without any comments interrupting, it will be registered as one edit.
"any comments interrupting" means that if you edit and then someone comments and then you edit again, it will be counted as two edits
So...can anyone define mathematics for me?
Maybe start with logical thinking and modify it till it fits? ;)
15:48
Hi chat
@MikeMiller Am confused about something, and could use someone with level of expertise
I can give it a shot
mmkay. It's probably silly, but here goes
I'm trying to reconcile two seemingly-contradictory statements re: holomorphic 1-forms on a Riemann surface
On one hand, I see statements like "Given a compact Riemann surface R of genus g, we have a basis $\{\omega_1,\cdots,\omega_g\}$ of holomorphic one-forms on R"
So that'd be a $g$-dimensional space.
Actually, I think I see where this is going. But let me proceed.
On the other hand, a genus-g surface would have homology group $\mathbb{Z}^{2g}$
So that's 2g-dimensional.
That seems like a mismatch. Shouldn't they have the same dimensionality by de Rham? @MikeMiller
probably that's too naive, but i would've expected there to be 2g holomorphic one-forms to match the homology.
In the sequence with the Bockstein homomorphism, I just computed that the map $H_n(X;\mathbb{Z}) \to H_n(X;\mathbb{Z})$ induced by the multiplikation $\mathbb{Z} \to \mathbb{Z}, x \mapsto k\cdot x$ is just multiplication with $k$, is that correct?
16:05
@MikeMiller Does the question make sense? I'm probably revealing my lack of sophistication, but it seems half as large as it should be
I think I see the resolution, actually: A smooth 1-form can be either holomorphic or antiholomorphic
so it's the distinction between the de Rham cohomology and the Dolbeaut cohomologies
(that would leave me confused for other reasons, though)
@Semiclassical Yeah, you just caught it. There is a splitting of de Rham groups into Dolbeaut groups, and holomorphic forms live in H^{1,0}. There is a duality which says that the dimensions of H^{p,q} and H^{q,p} are the same.
(I was showering)
16:40
@Semiclassical I don't think the TA really graded that SHO-Mehler kernel problem
@0celo7 Oh?
Nothing was checked, he usually checks off stuff
I got full credit
@MikeMiller What I'm confused about then becomes why, in a certain context, it seems like one needs to consider the de Rham group H^1 instead of the Dolbeaut group H^{1,0}
going to have to wrestle with that myself.
Well, one is harmonic forms, one is holomorphic forms
16:46
@Semiclassical: far from true that a smooth form must be holo or antiholo!
Hmm, chat on phone vastly better!
How do I understand Hodge decomposition, then? I'm probably saying something silly
Every closed form is cohomologous to a sum
@SemiC He's picking on the fact that a (1,0)-form is just one that has a holomorphic part (dz) and (0,1) anti holomorphic part (dzbar)
smooth = exact holo + exact anti-holo + harmonic?
We're doing only 1-forms?
16:50
yeah
@MikeMiller clarify? that didn't parse for me
No, there's a big difference between all smooth forms and closed ones, first.
(1,0)-forms look like fdz
but are not the same as holomorphic
I can write any smooth function times $dz$ and it won't be a holo form unless f is holo!
ah. so you're allowing f(z,w) dz (writing w=z-bar for convenience)
@TedShifrin Hi
16:53
hi @Tobias.
I think I've mixed myself up, then
Closed is a crucial ingredient !
hmm
Maybe I should say where this is coming from.
Unless we're doing the general Hodge decomposition of smooth. But that's exact + coexact + harmonic, all smooth.
Suppose I take the 1-form dx/y where $y^2=(x-e_1)(x-e_2)(x-e_3)$
that's a holomorphic (1,0)-form, right? (x,y in here are of course not real/imaginary parts)
16:56
Yes.
okay. now suppose I restrict to the case $y^2=x(x-1)(x-u)$ (I'm picking $u$ for the label of the moduli quite arbitrarily)
I hope my voice holds up for giving the seminar tomorrow
Almost done preparing it now
Dead voice?
I can then differentiate with respect to $u$ to get $\partial_u(dx/y)=(x^2-x)(dx/2y)$
@TedShifrin It has had some symptoms of it during the day (not that flying helps with that sort of thing)
and I have had a bad cold all week, which often ends with me completely losing my voice for a day or two

« first day (2283 days earlier)      last day (3033 days later) »