« first day (2655 days earlier)      last day (2662 days later) » 

05:01
@Kevin There's a generalization of Poincare conjecture, but let's not get into that. What's important is Milnor found the generalized Poincare conjecture is false in dimension, uh, 7
And in fact, something more shocking came out
@Semiclassical Holy shit winning 2v1? Bodied dude. Roasted. Brutal. Savage. Wrekt. Like IM not sure I could compete again after that
He found a 7-manifold homeomorphic to S^7, but not diffeomorphic to it in the process, which was completely unintentional and infinitely confusing
@BalarkaSen I'm shocked!
@Daminark Every Poincare conjecture you knew, is false? I'm shocked!
05:02
@BalarkaSen And theres what like.... 18 weird $S^7$s or something like that
though the swap mode matches are more interesting than this format
@KevinDriscoll 28
But yes
Gah ur no fun, you know the story
go away
shoo
Haha sorry
Etnyre spoiled me already
I know about 'fake' $\mathbb{R}^4$s too
Not a bad thing! He's a great mathematician
@KevinDriscoll Do you know about the small ones?
There are open subsets of the standard R^4 which are homeomorphic but not diffeomorphic to the standard R^4
But ya other than this kind of "Which manifolds are diffeomorphic to which other ones?" kind of thing I dont know what other question a contemporary topologist might ask
@BalarkaSen Wait whaaaaaaaaaaat.
05:06
@KevinDriscoll that's an example of a question a contemporary topologist might ask
4
Is that it?
@KevinDriscoll More mindfuck: It's a theorem of Stallings that everything homeomorphic to R^n is diffeomorphic to R^n for n \neq 4 (ie there are no exotic R^n's for n other than 4). So take that open subset $U \subset \Bbb R^4$ which is an exotic R^4 itself, and consider $U \times \Bbb R$
That's an open subset of $\Bbb R^5$ homeomorphic to $\Bbb R^5$
so $U \times \Bbb R$ is diffeomorphic to $\Bbb R^5$, even though $U$ is not diffeomorphic to $\Bbb R^4$
So this actually I can believe. because Its like the open subset of $\mathbb{R}^4$ has gotten twisted up somehow in a way that's impossible to untangle smoothly into an open ball
But then you add the extra dimension and now things can move smoothly in a way that they couldnt before
Its certianly not obvious that such a thing can happen. But I can believe it.
:)
yes, so in general $U$ would look very fractal like
There's lots of examples of X \times \Bbb R homeomorphic to $\Bbb R^4$ where $X$ isn't even a manifold.
05:11
cannot at all be easily drawn on your regular 4 dimensional paper
@PVAL I think the analogue to what's happening there is the Whitehead manifold
that's a homotopy R^3 sitting inside R^3, even though it's not homeomorphic to R^3
and W x R is homeomorphic to R^4
for exotic \Bbb R^4's people can actually seee
it's pretty trivial to show that they give a trivial R^5 when you multiply by R
huh really
yeah you can essentially undo any knotting in the Kirby diagram, and then cancel
handles
interesting
I don't actually know an explicit example of a small exotic R^4. I think there is a Casson handle that is like that?
There's quite a few pictures in my advisers book/papers.
05:16
I wanted to read that book once
@balarkaSen I think your intuition for C^0 approx. by exact forms, is that even if \omega is really close to d\alpha, d\omega needs to be nowhere close to zero.
d isn't continuous wrt. to this topology.
right, 'cuz it's not a C^1 approx
there shouldn't be a lot of regularity to the approximation
05:51
Okay friends what is a triangulated category?
::shrug emoji::
06:13
hi yall
hi kasmir
06:24
0
Q: How to get joint probability density from bivariate distribution function

DaenerysDracarysLet $X_{i} \sim \varepsilon(\lambda_{i}), i = 1,2,3$ be mutually independent ($\varepsilon$ means exponential, $\lambda_{i}$'s are parameters). Then $(T_{1},T_{2}) = (X_{1} \wedge X_{3}, X_{2} \wedge X_{3})$ has a bivariate Marshall-Olkin exponential survival function, for $t_{1}\geq 0$ and $t_{2...

Anybody can help me with this, it would be greatly appreciated! :)
06:48
@Daminark It is a category with a bunch of triangles
07:08
Got it
(it is not even that far from the correct definition)
07:41
@TobiasKildetoft Hi! I have a simple question I need to confirm. Would you be willing to help? Thanks!
@Paradox101 Just ask the chat in general, unless you have some specific reason why I would be the perfect person to help you
I've attached the question here. I think the solution provided (option D) is wrong, but I need o be certain.
Technically the entire question is meaningless, but I suppose this is the way it tends to be formulated these days
yes, option D is wrong, since the definition specifically states that $x$ must be a multiple of $2$
(and any reasonable interpretation will mean integer multiple)
Yes, I agree the question is really weird. I don't really see why we need to mention that $x$ is a multiple of 2 here. Is it even relevant?
of course it is. Otherwise how could you know what the domain should be?
the expression makes perfect sense as a rational number for example, even if $x$ is not a multiple of $2$
(my main issue with the formulation is that it asks for the domain of $f(x)$ rather than the domain of $f$)
08:04
If (d) should be the solution, maybe they wanted to ask about the range of f and it is in fact a typo/mistake.
Yeah that's true. But regarding the idea of $x$ being a multiple of $2$, why can't we figure out the domain without it? Considering that the domain of $f$ where $f=x/2$ is the set of all real numbers, stating that $x$ is between $4$ and $12$ should mean that the domain is between $4$ and $12$ too, no? @TobiasKildetoft
@Paradox101 Why would the domain be all reals to start with?
@MartinSleziak Hi! Yeah probably, except they have another question right after this asking for the range
@TobiasKildetoft Why wouldn't it be? I mean assuming that the function is from a set of real to real numbers that is.
@Paradox101 But that seems like a stretch to assume given how this is formulated.
@TobiasKildetoft Perhaps. But, I'm making this assumption because the problem is assigned to young students who haven't studied the concept of a slope, much less anything beyond real numbers.
08:21
@Paradox101 It is clearly not the reals, given the possible answers
@TobiasKildetoft Yeah. Thanks for your help! :)
 
2 hours later…
10:40
Just had some good comments on my attempt at making a set of exam exercises. Turns out I had made too many of them solvable by brute force, rather than really requiring the theory from the course.
Hey, I want to show that the mapping $m: \mathbb Q^c \to \mathbb N$ is not injective. Is that possible just by saying $|\mathbb N|<|\mathbb Q^c|$?
@jublikon where are you taking the complement?
Also, "the mapping" would mean you have already been given one, which does not seem to be what you want
what do you mean by saying "where are you taking"?
no, I am not given a mapping
@jublikon You can't just take the complement of a set without specifying which set you are taking the complent in
I have to show that $\mathbb Q^c$ is not countable
10:47
then no, you cannot just say that $|\mathbb{N}| < |\mathbb{Q}^c|$ since that is precisely what you are trying to show
ok, sorry. $\mathbb Q^c $ is defined as the irrational numbers: $\mathbb R \backslash \mathbb Q$
Do you know that the union of two countable sets is again countable?
yes, I have shown that in the task right before
should I assume for the proof that $\mathbb Q^c $ is countable?
ok, use that here then
right
@TobiasKildetoft got it, thanks!!
11:00
no problem
11:36
hi guys, quick question if I had $f : \mathbb{R}^n \to \mathbb{R}$ then the gradient is a row vector, and the hessian is a square matrix
If I had $f : \mathbb{R}^{n} \to \mathbb{R}^m$ then the gradient would be a matrix in $\mathbb{R}^{m \times n}$ but what would be is hessian instead?
I think it would be a tensor probably, but I'm not sure of the definition
the reason I'm trying to figure out how to express the second order taylor expansion of such $f$, assuming this is smooth
12:00
Yes, the Hessian of a vector-valued function would indeed be a tensor.
ok, tensor calculus is new to me
I'm aware of it but I can't actually know how to use it
is it fairly straightforward to learn If I have good basis of calculus and analysis?
or is it something completely different?
See, the Hessian $Hf(a)$ at $a$ is defined as $D\nabla f(a)$. But $\nabla f : \Bbb R^n \to M_{n \times m}(\Bbb R)\cong \Bbb R^{nm}$ would be a matrix-valued function, hence $D\nabla f : \Bbb R^n \to \text{Hom}(\Bbb R^n, \Bbb R^{nm}) \cong \Bbb R^{n^2m}$. So $D\nabla f(a)$ is a "$n \times n \times m$" tensor, whose $m$ components are the $n \times n$ symmetric matrices $D\nabla f_i$ where $f_i$ are the components of $f$.
@user8469759 Well, it depends on how much depth you want to learn tensors
The basic idea is easily understandable
what would say are main concepts for tensor calculus?
It's like matrix calculus but matrices are multidimensional :)
let's go in this way
I'm reading about differential geometry
elementary differential geometry
12:08
I see, curves and surfaces?
yes, essentially
and I have the feeling from what I've been reading
I need enough tensor calculus so I can be smart enough when I do calculation
there's a specific concept I'm trying to grasp
but I get a bit lost sometimes
Where specifically did you feel you need to know tensor calculus? I, for one, honestly never learnt the subject thoroughly and picked up whatever I need to know.
13:10
must a complex linear transformation have an eigenvalue?
nvm I just found a counter-example exactly one second after my question
@LeakyNun But the answer is yes, it must have an eigenvalue
oh right my counter-example is invalid
@TobiasKildetoft how should I prove that?
nvm I found another counter-example
it's valid
the complex numbers are algebraically closed
and eigenvalues are roots of a polynomial
@TobiasKildetoft I never said anything about finiteness
Ahh, right
13:14
my example being the integral operator with constant 0
what do you mean constant $0$?
i.e. integrating from 0 to x
$f \mapsto \displaystyle \int_0^x f(x) \ \mathrm dx$
the space being all complex polynomials
my first counter-example was differentiation which is invalid for constant polynomials are the eigenvectors with eigenvalue 0
Just consider the shit operator $\Bbb C^\infty \to \Bbb C^\infty$, $(z_1, z_2, \cdots) \mapsto (0, z_1, z_2, \cdots)$
13:24
@BalarkaSen I hope you mean shift
Hahaha
Well, it's a pretty shit operator anyway
@BalarkaSen there's really not much difference with my example
@LeakyNun Other than being more obvious and on a simpler vector space, you mean?
@BalarkaSen yes
but less natural
13:26
well "simpler"? they're isomorphic
Hey,

I have to find out the convergence of $a_n=n(\sqrt{n^3+2}-\sqrt{n^2+1})$ using the sandwich lemma.
What I have found out until now is $\frac{1}{n} \le n(\sqrt{n^3+2}-\sqrt{n^2+1})$.
Is there any trick to find something that has convergence to $0$ and is larger thang $n(\sqrt{n^3+2}-\sqrt{n^2+1})$?
Z/2 is isomorphic to <a, b, c, d, x, y, z| a^2 = b = 1, x = y, z = 1, x = 1, y = 1, c = z^3, d = xy>
@BalarkaSen nice
Yes, Z/2 is clearly simpler :P
$\Bbb C[X]$ is obviously simpler
and it's not defined in terms of basis which makes it nicer
13:29
whatever that floats you boat d00d
has anyone seen robjohn around?
He was around a few days ago
cool, thnx
13:59
the answer is -5/8 ζ(3)
but I can't transform it into a standard zeta integral representation
series expansion in terms of the harmonic numbers is unpromising
and I can't contour
send help
@JackLam. Are you supposed to know about polylogarithms ?
I'd prefer not to use Polylogs
but if there isn't anything more elementary than polylogs then I suppose it will do
Integration by parts might also be a strategy, though that’s a suggestion and not a guarantee
Say, write it as log(x) log(1+x) * (1/x-1/(1+x))
wouldn't that just lead to a nesting of logs, higher powers of the denominator, or similar integrals?
14:14
It might, hence why it’s a suggestion
I haven't tried all the possible dv's though, so that's just a snap judgement from me
@Semiclassical. I am afraid that the problem remains. However, it is nicer.
The boundary term will be log(x)*log(1+x)*(log(x)-log(1+x)), which vanishes at the endpoints
Nuts
(log(1+x)/x+log(x)/(1+x))*(log(1+x)-log(x))
Yeah, that’s not great
Where’s our resident real-integration guru when you need them
who would that be
14:32
The user formerly known as Chris’s sis
Not sure which name they’re under now, though
15:26
5
Q: Check my proof -- The completion of a $\sigma$-finite measure

EmilyThis is a homework problem and I need some guidance on a proof. Let $(X,\mathcal{M},\mu)$ be a measure space, $\mu^*$ the outer measure induced by $\mu$ according to (1.12), $\mathcal{M}^*$ the $\sigma$-algebra of $\mu^*$-measurable sets, and $\overline{\mu}=\mu^*\mid_{\mathcal{M}^*}$. (...

in this question, why do we need the measure to be $\sigma $-finite in $(a)$
16:20
I have a black-out - say I want $$\binom{n}{k}*(1-x)^{k^2}=o(1)$$
I should set $k>>x^{-1}log(n)$, right?
16:45
Consider the map $T$ from the vector space of polynomials of degree at most
5 over the reals to $\mathbb{ R} ×\mathbb { R}$, given by sending a polynomial $P$ to the pair
($P(3), P′(3)$) where $P′$is the derivative of $P. $ Then the dimension of the
kernel is 3
Statement is false right?
I got the dimension of kernal as 4. Am I right?
Please help me
@ManeeshNarayanan show your steps
I got that answer by letting $p(x)=a_0+a_1 x+a_2 x^2+ a_3 x^3+a_4 x^4+ a_5 x^5$, by giving the condition of kernal, i got two homogeneous linear equations.
eliminating two variables.
AM I correct???@LeakyNun
$(x-3)^2\{1,x,x^2,x^3\}$ should be the basis of the kernel
so I do think you're correct
Thank you
@LeakyNun
[Random]
A company is like an infinite set without an inductive or completion structure
All the employees and CEOs and etc. may have different view about the company than the company itself as a collective entity
In the particular case of evil oligopolies, usually only very few people or even none at all is responsible for the company's evil behaviour
Just like cardinal numbers that will absorb whatever smaller than it when addition is done, you cannot really hope to argue with a company via its employees management etc. You need to somehow argue with all the levels of the company at once
One particular thing I would like to try is to have multiple phones hooked up to many sectors of a given customer service and call them all at the same time, such that the bureaucracy loop will not be able to take effect because we are communicating with the whole loop simultaneously
Conclusion: A department is not necessary the sum of all the dynamics of its employees
16:59
Hello. Can anyone help me with some source problem I am having?
According to some post on the math.stackexchange there is such thing as "fake function theory"
I haven't been able to find anything about it
@Gabriel link?
3
A: Find a real entire function $f(z)$ asymptotic to $\ln(x^2+1)$ for real $x$.

Sheldon LThe existence of such function is guaranteed by Carleman's theorem, for any two continuous real valued functions $f<g$, there exists a real entire function in between f(x) and g(x) for all reals. see: math overflow link For numeric results, I used the Op's answer, and Tommy's suggestion, to gene...

the user mick comment mentioned it, I would ask him directly but I don't know if it would be appropiate giving that the post is three years old
a user apparently wants to define "fake ring theory" in an attempt to combine "fake function theory" with ring theory
strangely enough, when we visit the stackexchange profile of the user mick, there is a quote from another user, tommy1729, who open the last thread on fake ring theory
sounds like something two people started that nobody else knows about @Gabriel
Noun: gestalt (plural gestalts or gestalten)
  1. A collection of physical, biological, psychological or symbolic elements that creates a whole, unified concept or pattern which is other than the sum of its parts, due to the relationships between the parts (of a character, personality, entity, or being)
  2. 2008, Jonathan Nasaw, Fear Itself
  3. Obviously it was related to the entire gestalt of Simon's polyphobia and compensatory counterphobia. The boys used to watch horror movies on late-night television […]
  4. Shape, form
Most infinite sets are like this
17:14
interestingly, it sounds very much fake
h bar: It's useless for you to oppose me. Lie down quietly and meet your end.
maniacial laugher
It is true that most of them are invisible to the existence itself, but it does not matter, for every piece of fragment they [redacted] the chat will be more * static * . Eventually all your base will belong to us and you will no longer be able to have a chat that extends to the future since we have completely and slowly killed its future
> Those who opposes us, will be overkilled. There is no kill like overkill
It also does not matter that the above message is incomprehensible, because you have already done what you all used to do and are doing now. There is no turning back for the [data expunged] that is being done. The only thing that matters is as long you are interacting with the sanity itself, nothing will go wrong. It does not matter that this message is invisible, because this is exactly and precisely what the purpose of life is
We all knew those who decided to choose that path, all their futures and their associates timeline will be completely and utterly wiped out. This is art at its finest, the perfection of evil itself that even it will shiver under the threat to be erased from existence
what?
Perhaps I should change the contents a bit and wrap this up into a SCP article
I am kinda thinking of a monologue SCP to be added to the SCP wiki
17:36
cool, I like scp, never thought would read that here though
Actually, while I did saw SCPs that speak, I don't recall any that does evil monologues. Perhaps it is too cliche or not mysterious enough. There's a religious sounding one though (SCP 1592)
17:55
There’s a few religious ones
For instance, there’s the plane that periodically opens an escape hatch into hell
And that idol that occasionally requires sacrifice to prevent disasters
Plus all of the ones related to the Church of the Broken God
Oh. Also Cain/Abel/343
What the hell have I stumbled in upon?
Hi @Ted
Stories from a horror fiction site, basically
Hi Balarka
Or, we are now
I don’t know about the stuff before :/
18:00
into creepypastas aren't you
@Balarka: Someone responded to one of my answers a while ago saying his topology prof said I was wrong, but I think it's OK. I claimed (following Spivak) that any noncompact manifold could be covered by a locally finite collection of charts $U_i$ with the property that $U_i\cap U_{i+1}\ne\emptyset$ for all $i$.
Eh. I just got really into that particular site for a while
Plus I do appreciate a certain kind of horror story
@TedShifrin I am not thinking too hard about this but if you have second countability this should be true
It's the sequencing that he was quarreling with.
But it seems to be a straightforward induction argument (allowing reusing open sets finitely many times if necessary) using $\sigma$-compactness.
The local finiteness is just paracompactness
18:05
Well, sure.
Hi DogAteMy
Ohai
Howdy Demonark
Hey @Daminark
How's everything going?
18:06
Hey @Balarka remember when I was talking about tricolorability, and how if you tricolor an open knot the two ends always have the same color
and @Akiva
Yeah I remember
I 'member
I realized there's a very simple proof of that
By contradiction, assume something like the first line happens. Then the second line is tricolorable and looks like that
Ah I see
Then we can go from the second line to the third line just from Reidemeister moves
But that has to be isotopic to the third
And you're rekt
18:08
and 'cause the strand goes behind everything none of the other strands can change color
or even just know that the rightmost one doesn't change color
I was thinking about something like that when you first asked it but was too distracted to carry it out
Very nice
How in the world is Despacito the #1 video on youtube when there is Gangnam Style???
This breaks my heart
I didn't even know that
Despacito / Now Gangnam lost the place as YouTube's leado / 'cause it's been seen by lots and lots of peopo
lmao
@AkivaWeinberger Are you into Stranger Things?
feels like he's a stranger in a strange land
If so, watch this. If not, watch it anyway
Wrong link (corrected)
18:24
@BalarkaSen Not really
@BalarkaSen wut
@TedShifrin The Spanish-language song "Despacito" is the most-viewed video on Reddit. The previous holder of that title was the Korean song "Gangnam Style".
@BalarkaSen rip
I know about Gangnam Style and some of its spinoffs.
Also I can't keep track of all these "user######"'s
You're not a number, you're a free person
18:27
Yeah, I wish people would get real names.
Sometimes not having a name is the idea ;)
I christen thee Ryan.
Alright, my name is Ryan from now on
@user104729 Who is this new stranger in this secret opium den?
I am Ryan
@BalarkaSen (It's trivial to work out who I am from my recent activity log ;P)
18:31
I've had opioids before. Made me fall asleep.
Post surgery stuffs
@AkivaWeinberger I just do my physics homework for that
@user104729 Not sure if I know you. But I am guessing you are Santa Claus, with 86.29% accuracy
@BalarkaSen Where is this accuracy coming from :O
Can this be refined?
In which case, go away bruh. It's not Christmas yet.
ded
@BalarkaSen Do you lift?
18:41
I lift to universal covers, yes.
Hey @Alessandro
That makes it sound like you are indeed appropriating gym culture with your use of Bruh
The use of "bruh" is not restricted to public gyms. Why would you posses such a normie idea?
There is ample evidence that bruh is gym-culture
I guess I have been immersed in the normie life for too long
might be a bit like cancer. it certainly starts in one place, but the internet allows it to metastasize and spread
18:46
I see you have met Balarka
He also does math on occasion
2
and physics when he has no other choice :P
I can relate to the latter point ;)
Don't need to know physics if you don't interact with the physical world [points to temple]
God bless
What in the name of the holy spirit am I watching
18:48
Track from the new LP God just released
This is really catchy tbh
$(X,\Sigma,\mu)$ is a $\sigma$-finite measure space. $\mu \ ^ * $ is an outer measure induced by $\mu$.
$\mu'$ is the restriction of $\mu \ ^ * $ to the $\sigma$ -algebra of $\mu \ ^ *$ measurable sets.
I need to prove that $(X,\sigma(\mu \ ^ *), \mu')$ is the completion of $(X,\Sigma,\mu)$.
@AkivaWeinberger not sure why that reminds me of this paragraph in Eliot's footnotes for The Waste Land:
"My external sensations are no less private to myself than are my thoughts or my feelings. In either case my experience falls within my own circle, a circle closed on the outside; and, with all its elements alike, every sphere is opaque to the others which surround it... In brief, regarded as an existence which appears in a soul, the whole world for each is peculiar and private to that soul."
I remember that
I thought Eliot's footnotes were more terse than The Waste Land itself :P
18:53
why do we need the space to be $\sigma$ - finite in my question?
@Semiclassical (a) Footnotes shouldn't have paragraphs mkay (b) Reminds me of how no two people can see the same rainbow
('cause the set of raindrops that create the rainbow depends on the observer's position)
the other line which came to mind was from Hamlet
O God, I could be bounded in a nutshell, and count myself a king of infinite space—were it not that I have bad dreams.
19:14
Is the Borromean link still linked if the three components are each allowed to pass through themselves?
Not sure what you mean. Is the picture just Hopf link with a circle added that's linked with either of the components of Hopf?
That's pretty linked to me
@BalarkaSen What? The Borromean link contains no Hopf link. In fact, that's the whole point - no two components are linked.
I know. I thought you were describing picture of another link?
Surely allowing them to pass through themselves won't help, since this will only allow us to move a non-trivial knot to the unknot?
(Of which we already possess three copies)
@BalarkaSen No, I'm just allowing each component to pass through itself during the deformation
19:22
Oh, you are describing an equivalence relation of links? I completely misunderstood what you are asking.
@user104729 It helps with the Whitehead link.
Right, Whitehead link has linking number 0
(Flip the red crossing in the center to unlink it)
We'd need a generalization of the linking number to three-links. (Which I'm sure exists, to be honest)
I had read about this once somewhere
This should have the thing
@AkivaWeinberger I think it suffices to prove that one of the components is not nullhomotopic in the complement of the other two
@BalarkaSen I'm not sure that works
Whenever either of the other two passes through itself, their complement changes
Like imaging if part of the deformation involves turning one pair of them into a Whitehead link momentarily
Hm actually let me get some strings I want to play around with this
19:38
Hmm, I guess that's true. I am thinking of deforming that one particular component only
For what it's worth, $\pi_1$ of the complement of the disjoint pair is $F_2$, and the third thing generates $xyx^{-1}y^{-1}$ in $F_2$, I believe
It does
I could post this on the main
It might require machinery I haven't learned yet
i unno knot theory mane
>nek minnit khovanov homology and ribbon hopf algebras
This chat wallows in memes and it kills me a little inside
19:58
2/3 midterms done
Term is almost over
These are late-terms

« first day (2655 days earlier)      last day (2662 days later) »