« first day (2281 days earlier)      last day (3035 days later) » 

00:00
I am, however, going to try to understand Bryant's paper on Chern's discoveries regarding $G_2$ and the non-existence of complex structures on $S^6$. That's differential systems stuff that I can follow :)
@TedShifrin Do you have a link for that paper?
It is one of the Atiyah references. I had downloaded it ages ago, though.
Thats good enough I can find it
You'll have to learn a lot :)
I have friends now
That are also interested in G2
00:02
G_2 is a good group.
So its something to talk about maybe
yeah, but this is the Cartan game with lots and lots of differential forms
@Ali It's not quite the same as the way they think about it.
I'm not used to $G_2$ (sadly), but ...
I can understand it in terms of the group preserving ...
@TedShifrin If you end up with a paper settling this question, remember to acknowledge me as the one who showed you that paper.
00:04
@TedShifrin Your knowledge beats mine however
OK, @PVAL. I should be on the lookout for anything Bryant writes, but that's almost a full-time job and I'm retired. :)
If someone asked me the difference between G2 and SO(7) I don't think I would be able to say much
@Ali, you've thought more about $G_2$ than I have. But I've thought a lot more about Cartan-Kähler theory and such things ... :)
@PVAL: I still remember losing my first job prospect to Robert Bryant. He became a super-star almost overnight.
From July this year to 2027, the barycenter of the solar system is not inside the sun
But we're good friends.
00:05
So we're not really orbiting the sun
I find this interesting
That's not shocking, DogAteMy.
Wait a minute. Barycenter? Not center of gravity?
This is despite how the sun is >99% the mass of the solar system.
@AkivaWeinberger So we are a binary system with the sun and the rest of the solar system?
Are they not the same? Center of gravity
I only know the word barycenter in a totally non-physical context.
Nothing to do with masses, in particular.
00:07
@AliCaglayan I suppose only Jupiter is really "not orbiting the sun"ā€¦
What about Pluto? :D
@TedShifrin Too busy as a binary dwarf planet system
Wait, no
I think it applies to Earth too
because we're pulled by Jupiter
(the only other mass that matters, except for maybe Saturn)
00:08
We are pulled by everything
The rest are really light
Or rather, have little mass
in comparison
woo
space
Well you could argue that nothing orbits anything
Its all nonsense really
Rather nihilist
00:11
We are all just following a geodesic
I'm iconoclastic. I never follow geodesics.
Iconoclastic sounds like a word at a spelling bee
Are you insulting me again, @Ali?
Oddly enough, that was one of the vocab words I learned today in class
Iconoclastic
00:12
@TedShifrin It has more letters than I have fingers I am merely uninterested
my iphone offers šŸ˜’ as an autocomplete option for boo
LOL ... I aim to serve, DogAteMy
smacks @Ali
haha fooled you I am balarka in disguise
Good disguise.
Don't you hate it when all possible geodesics from your position to the future lead to the center of a black hole?
00:14
@AkivaWeinberger prove it
I'd rather not
Geodesics are quite hard to compute.
I suppose it's possible I enter a black hole and find I rather like it, but I doubt it
But will the black hole like it?
I mean if hawking radiation is correct then this is all, in technical terms, nonsense?
"was it good for you too?"
00:17
That can't be @Alessandro. It's hours past his bedtime.
I don't even know what bed or time is anymore
Oh actually while @TedShifrin is here I can annoy him and ask him to help me with my homework
Why does everyone derive pleasure from annoying me?
I don't.
for intervals is infinity commonly a ( or [ ? I see a lot of (-infinity,infinity) but surely you would want to include the "infinitely last number" so it should be [-infinity,infinity]. Does it make much difference when we are talking about infinity?
@WDUK What is infinity?
00:20
$test$
@WDUK: Good luck finding that number.
regardless of whether i can find the number
why do people use ( and not [
there is no real number called infinity. there is no last real number.
Because it's a symbol ... there is no number.
pre-emptive hi
00:21
Writing $(a,\infty)$ is just a shorthand for writing all numbers greater than $a$.
well then if anything neither ( or [ is correct
If we have a connected open set U (subset R) then a continuous map z to S1, then we can find a map t such that z(x) = exp(i t(x)).
well, there are definitions where infinity is a point no?
retroactive hi @tern
I need to find a counter example for the case when U is not connected
00:22
yes, @Ali, but that's surprisingly nontrivial to prove.
@TedShifrin i get that but wouldn't (a, inf] be more correct =/
So that a t cannot be found
@saturatedexpo there are situations where infinity refers to a point in a topological space, yes. the riemann sphere has one for instance. but not R.
@WDUK No, since it doesn't include infinity.
NOOO, @WDUK, because that would include infinity and there is no such thing.
00:23
@arctictern what you doin lately
gets tired of these stubborn arguments
@WDUK As a definition, $[a,b]$ is the set of numbers from $a$ to $b$ inclusive, $[a,\infty)$ is the set of numbers greater than or equal to $a$, etc. It's a lot of casework, which is annoying, but it's the easiest way to get to a definition.
@TedShifrin So basically what I am trying to say is when U is not connected why does it break
(If you're wondering what the dollar signs are, see if you can get LaTeX ā€” @WDUK
ā€” should be on the right.
@MikeMiller tried thinking about clifford algebras again. now understand how to generate the table of Cliff(p,q)s, mod 8 periodicity.
00:24
Not LaTeX, MathJax. Sorry.)
@Ali: Doesn't it apply on every connected component?
@TedShifrin I should have probably mentioned that these functions are smooth
What's the domain/range of $t$?
Irrelephant.
tonight been confused about super morita equivalence. will start "ideals, varieties and algorithms" textbooks starting tomorrow
I'm just worried he's gonna get confused when he discovers $\bar{\Bbb R}$
00:25
Ok let me whip out the exact question
one can use $[-\infty,\infty]$ in things like an extended real number system $\Bbb R\cup\{+\infty,-\infty\}$, but this is extended
@TedShifrin Irrelephant is memeworthy :D
I have bad memories of memes, but go ahead, @saturatedexpo.
I am proving mu is an inner product. Mu has the following condition: If vāˆˆVvāˆˆV satisfies Ī¼(v,w)=0Ī¼(v,w)=0 for all wāˆˆVwāˆˆV, then v=0vv=0v. Someone answered that this condition "tells you that the map has a trivial kernel. However, an inner product must be positive-definite, which may fail even if the kernel is reduced to the null vector." Can someone explain this?
Your typing has serious problems, @Aksel.
00:28
yeah just saw that, it was copy and pasted.....
Give an example of an open $U$ in $\Bbb R^2$ and a smooth function $w$ from $U$ to $S^1$ that is not the form of $\exp(i t)$ for a smooth $t$ from $U$ to $\Bbb R$
$\mu$ could also be negative-definite, @Aksel.
@Ali: Does it require $U$ open?
@TedShifrin I do not understand negative definite or how to show this...
00:29
@TedShifrin Yes sorry
> I have bad memories of memes
wut
@Aksel: Do you know what positive definite means?
@arctictern sounds fun, but you're turning into @AndrewT
Tern: Do you not remember some of Hippa's mean memes?
LOL @MikeM
is the hole study of math at the university proving theorems?
00:30
I think I have something somewhere on Hannibal using the elephant of surprise
@TedShifrin Are those the ones that made you regret your video lectures
@TedShifrin not entirely
depends on the fields one studies, @saturatedexpo ... and that's "whole," not "hole" :P
@Ali: I never regretted my lectures ... I regretted Hippa.
But now I miss him, actually.
2 days ago, by DHMO
> We often hear that mathematics consists mainly of "proving theorems." Is a writer's job mainly that of "writing sentences?" - Gian-Carlo Rota
was looking for that quote
00:31
@TedShifrin depends how you look at it, from the top or the bottom :D (hole/whole)
So I am trying to show that no such smooth log function exists
Well, learn the definitions, @Aksel. Change the sign from + to - and it's negative-definite.
Therefore t = log . w
Will never exist
But I am very much stuck
@Ali: But if you can do it on a connected open set, you can do it on any component of any open set ... and smoothness is local.
oooohh
Wait
The question was wordered horribly
00:33
Damn sloppy professors.
Worded
I have to give an example of an open U and some smooth w where this doesn't hold
OK I can probably do that
That's no different, @Ali. Get some sleep!
I can't @Ted
I woke up at 4 pm today
You're screwed
But it still is correct, regardless.
currently i have begun with analysis(I) and linear algebra(I), seems infested with proofs. I guess my real question is, can those proof techniques be applied to another field? (e.g. proofs by computers, or pointing to fallacies in textes)
00:34
Does Kazumasa Kuwada ring a bell?
um, no
He is hosting an after party for the geometry seminar
@saturatedexpo: what country are you in?
As in a seminar after
@TedShifrin Germany
00:36
Well, in Europe there's much more emphasis on theoretical mathematics than, say, in the US. There is, nevertheless, a lot of good applied mathematics. Take some courses in numerical analysis, cryptography, etc.
Statistics is really big here in the UK
Statistics is really quite separate from math(s) ...
But, yes, take probability and statistics and ...
isnt statistic already applied mathematics?
Well I mean there are a lot of masters students interested in stochastic processes here
thats what I meant by statistics probably
Anyhow, @Ali, I think your question has no answer.
00:38
on my university statistics is a duty anyways for bachelor.
Ohhhh, did you say $U\subset\Bbb R^2$, @Ali? Seriously?
I thought you said $\subset\Bbb R$ the whole time.
Am I just really bad at communicating
No, you had it the whole time. My apologies.
It can be connected. No problem.
Part (b) is your total clue.
my chatjax kinda doesnt work on firefox, do i have to install something besides clicking the link?
00:40
Part b is just because U is connected right
which is doable
I can prove this for when U is connected
NOOOO, @Ali. STAR-SHAPED is essential (well, not literally necessary)
Um, no, you can't
Part (c) is a counterexample to your proof.
I don't know how Firefox works; does it have bookmarks?
Your example in (c) can well be connected. He should have put connected in there.
ah now with the bookmark it works thanks @AliCaglayan
Sure it has bookmarks. Are you on a phone or tablet, @saturatedexpo? Then it's a challenge.
It's Akiva who gets thanked :P
00:42
oh^^
Thanks is transitive right?
Thanks Akiva
thanks @AkivaWeinberger hehe
:P
You're welcome.
did you made your master already @TedShifrin ? Did you already forgot alot what you did in the beginning?
00:44
He's a retired professor
Huh? @saturatedexpo?
@TedShifrin I am confused, are you sure the last bit needs to be connected.
This is how I see it. For a) I prove that this holds for a connected U. For b) I show that star-shaped means connected so it holds.
@Ali: If you can prove something about continuity/smoothness on a connected open set, you can prove it on any open set by doing it component by component. It does not need to be connected, but he should have said connected, yes.
NOOO ... @Ali. See, this is why you confuzled me. Part (a) is for intervals in $\Bbb R$. Part (b) and (c) move to $\Bbb R^2$.
Oh I have confused myself also
See, that's not smart :P
00:45
Reading the question is the hardest part of these problems
@TedShifrin I mean, for me this is really alot, not necessarily hard to understand. I wonder if later you just keep handwaving your way through with lemmas you dont know anymore.
Hint to you for (c): What's the simplest (open) set in $\Bbb R^2$ that is not star-shaped with respect to the origin?
No, @saturatedexpo. I didn't spend my life teaching/researching by waving my hands at things I didn't know.
Any open set that does not contain the origin?
Be more specific, @Ali.
$\Bbb R^2 \setminus \{ 0\}$
00:47
LOL, OK.
An open set whose closure contains $0$ but that does not itself
OK, @Ali, now you can do your homework.
Thank you @TedShifrin
I have to get back to packing for my trip.
Is a proof by induction flawed if you did not show that the assumption holds for a specific n? (i.e. n=1)
Yes.
You haven't proved that it holds for any $n$, in that case.
but can then induction be a tedious way to prove? because if you can't find that n you are screwed.
00:53
You only need 1 n
Usually induction is good at proving something you already know
ah, and then you usally already now for what n it holds
Well if you prove something holding for n holds for n+1 then thats great
give an example of an n and you have all natrual numbers you care about
is there something as n->n-1 or will this produce alot of fakeproofs?
Induction is like an infinite IOU. But you need somebody to pay the bill for the chain to resolve
@saturatedexpo You could do that but it wouldn't be useful for natrual numbers
Maybe over the integers
but n => n-1 is not the same as n-1=> n
so for a statement like "all integers have a predecessor" (altho this might be an axiom already)
01:02
Well its a useless statement
It may be true but it doesn't tell you anything about sucessors
01:27
0
Q: Area of triangle formed from equation of of its sides.

RamanujanThe area of triangle formed by the lines $$a_1x + b_1y + c_1z=0$$ $$a_2x + b_2y + c_2z=0$$ $$a_3x + b_3y + c_3z=0$$ Is $$\frac {\Delta^2}{2\lambda_1 \lambda_2 \lambda_3}$$ Where $$\Delta = \begin{vmatrix} a_1 & b_1 & c_1 \\ a_2 & a_2 & c_2 \\ a_3 & b_3 & c_3...

is there a simple reason that exacly 5 unique lagrangian points exist between 2 large bodies
@shaihorowitz I don't think there is a nicer reason than saying that if you have two bodies and you calculate the effective potential a third body would have due to gravity and centrifugal force then there are 5 stationary points
If you want to think of the symmetry of it they appear in a nice way.
L
L Body1 L Body 2 L
L
Those L's above and below are supposed to be above the middle L
They form sort of a star shape
.. L
L L L
.. L
that is a nice form based off symmetry but it leaves the question of why not the elipse with the body on the foci
What do you mean
it really intuitevly feels like the 0 energy points should either be continously connected or based off of the zeroes of a polynomial
like why not
...L
..LL
.LLL
LBLBL
01:46
@shaihorowitz So it doesn't look like there is a nicer reason
lol t his guy on vixra tried to prove P=NP
and he literally said that "N=1"
i dont think he realized P and N aren't variables lol
yeah thats a messy contour plot doesnt look to simplify nicely into a single fifth order equation
well the p=np proof ought to be interesting
let me know if he ever derives $P=-P$ - means not @meow-mix
@shaihorowitz well actually it does
look at equation 10
02:01
yeah but that only gives solutions to 3 points and 2 imaginary ones the other points come through considering an additional force
i typed Reduce[(x - y)/Abs[x - y]^3 + (x - z)/Abs[x - z]^3 == 0] into mathematica and now its broke
02:20
@Shai: How do you expect it to reduce it?
Long division
How do i prove: if n>8 is a perfect cube, then n-1 is not prime
what does "n is a perfect cube" mean?
there is integer k^3=n
so you're trying to prove k^3 - 1 is not prime, with k>2
can you factor k^3 - 1 at all?
yup.... got it now. TY
02:46
mmhmm
> Proof:$$\bigcup^{\infty}_{n=1}\left(\bigcap^{\infty}_{m=n}{X_m}\right)=\{x:x\not\in X_m \text{ for at most finite many } m\in\mathbb{N}\}$$

I tried this, with no great luck:

$$\bigcup^{\infty}_{n=1}\left(\bigcap^{\infty}_{m=n}{X_m}\right)=\bigcup^{\infty}_{n=3}\left(\bigcap^{\infty}_{m=n}{X_m}\right)\cup\bigcup^{\infty}_{n=2}\left(\bigcap^{\infty}_{m=n}{X_m}\right)\cup\bigcup^{\infty}_{n=1}\left(\bigcap^{\infty}_{m=n}{X_m}\right)\\=\bigcup^{\infty}_{n=3}\left(\bigcap^{\infty}_{m=n}{X_m}\right)\cup\bigcup^{\infty}_{n=2}\left(\bigcap^{\infty}_{m=n}{X_m}\right)\cup X_1=\bigcup^{\infty}_{n=
show every element of one set is in the other
using logic and words
also, proof is a noun, the verb is prove (/pedantry)
hmmok, I'll retry. Thanks for refreshing my English-skills. :)
how about: if n is odd, then $4^n-1$ is prime
so $n=2k$ for some integer $K$. then $4^n-1=4^{2k+1}-1 = 16^k4-1$. how do i show that's prime?
03:02
63 is prime?
@TedShifrin no but 57 is
ok. so that one's false (that was the first part of the question)
Thanks, PVAL ... You love correcting me these days :)
@arctictern I really don't know how I'm supposed to do that here. if I assume that $x\in\text{The \bigcup stuff}$ I have the same problem as before with the "finiteness". It's still infinitely many $X_1\cup X_2\cup\cdots$.
@NaCl $x\in \bigcup_{n=1}^\infty \bigcap_{m=n}^\infty X_m$ if and only if there exists an $n$ for which $x\in \bigcap_{m=n}^\infty X_m$ if and only if $x$ is in all but finitely man $X_m$s
03:06
@Ted I looked at a couple questions after thinking they had interesting titles on this site. They were both random philosophy about definitions. So it's a better use of my time than that.
@Jeff also, I tried to contradict, see $4^{2n-1}-1=p\cdot q$ for any $p,q$. Then $4^{2n-1}-1=\frac{4^{2n}}{4}-\frac{4}{4}=\frac{4^{2n}-4}{4}=pq$ which clearly holds, no contradiction
LOL ... Poor baby :)
@arctictern Why finitely many? You got $\bigcap^{\infty}_{m=n}$
@NaCl If $x\in \bigcap_{m=n}^\infty X_m$ then $x$ is in every $X_m$ except for possibly $X_1,\cdots,X_{n-1}$, which is finitely many $X_m$s. Conversely, if $x$ is in all but finitely man $X_m$, then let $X_{n-1}$ be the one it's not in with highest subscript, then $x$ is in $\bigcap_{m=n}^\infty X_m$.
ooooh!! I missed that, you're right. Thanks man! :)
03:44
The proof is not as trivial as I thought
@arctictern remember that set of functions I was asking about before?
I don't know for sure if they are one and the same
but it would appear a quality both sets have in common is that "no inverse exists anywhere"
so I guess one way to look at is that I looking for a branch of functions that "have no inverse anywhere"
04:15
Is there a linear transformation that fullfils
$T(1,0,0) = (1,1,0,0)$
$T(0,1,1) = (1,2,0,0)$ ?
@Maks Try sending $(0,1,0)$ to $(1,0,0,0)$ and $(0,0,1)$ to $(0,2,0,0)$.
Hi @Fargle
Hey @Balarka, what's crackin'?
@Fargle I need to find the equation of T
Not much; thinking a little bit about topology and geometry.
04:20
@Maks Recall that the columns of the matrix of $T$ are what the basis vectors get sent to by $T$.
@Fargle And how should I build the matrix ??
@Balarka: sounds fun. I've been doing some Spivak and Rudin exercises to help build my intuition.
@Maks Well, the first column is what you want $T(1,0,0)$ to be, so what would that be?
(1,1,0,0) ?
04:23
Exactly!
So the matrix is
Now the second and third columns are what you want $T(0,1,0)$ and $T(0,0,1)$ to be, respectively.
$ \begin{pmatrix}
1 & 1\\
1 & 2\\
0 & 0\\
0 & 0\
\end{pmatrix} $
Ohh wait
Why T(0,1,0)
and $T(0,0,1)$
Because I look for the basis vectors ?
Right.
But I dont know their value
04:25
You can choose any values so that the matrix will follow your constraint.
You know that $(0,1,1) = (0,1,0) + (0,0,1)$, so
You want $T(0,1,1) = T(0,1,0) + T(0,0,1)$.
So I choose any value I want ?
Like T(0,1,0) = (1,1,0,0) and T(0,0,1) = (0,1,0,0) ?
@Maks Any value that gives you $T(0,1,0) + T(0,0,1) = (1,2,0,0)$ as per your initial constraint.
$ \begin{pmatrix}
1 & 1 & 0\\
1 & 1 & 1\\
0 & 0 & 0\\
0 & 0 & 0\
\end{pmatrix} $
So that matrix is ok ?
I should think so, yeah.
I reduce
\begin{pmatrix}
1 & 1 & 0\\
0 & 0 & 1\\
0 & 0 & 0\\
0 & 0 & 0\
\end{pmatrix}
I end up with that
What now ?
x+y = x
z = y
I mean
$x+y = x$
$y = z$
$z = 0$
$w = 0$
Is that ok ?
04:31
Well, all you were looking for was a linear transformation, right?
But that doesnt work
T(x,y,z) = (x+y,z,0,0)
T(1,0,0) should be (1,1,0,0)
The RREF of a matrix is NOT equivalent to the original matrix.
They share some properties but don't do exactly the same thing.
So I dont have to reduce it ?
Right.
$x = x+y$
$y = x+y+z$
$z = 0$
$w = 0$
T(x,y,z) = (x+y,x+y+z,0,0)
T(1,0,0) = (1,1,0,0)
T(0,1,1) = (1,2,0,0)
04:33
There you are.
Thz man !
No problem!
It doesnt have only one answer right ?
Cause I can pick more than one combination of vectors
Any value that gives you T(0,1,0)+T(0,0,1)=(1,2,0,0)T(0,1,0)+T(0,0,1)=(1,2,0,0) as per your initial constraint. there isnt only one
04:51
@Maks Correct.
There are infinitely many answers.
And is there another way to get those two vectors ?
That's not guessing ?
Like creating a matrix or some kind of equation ?
 
3 hours later…
07:56
Iā€™m trying to prove that there exists an ultrafilter on the natural numbers which contains the even numbers. Can anyone help?
08:10
anyone know the amazing identity $\cot(x)=\sum_{n=-\infty}^\infty \frac{1}{x-n\pi}$?
08:32
@Sophie You can derive the identity by differentiating the infinite sin product
that's exactly how I found it
it's interesting because you get $\zeta(2n)$ in the power series, then those are easy to compute for all n
@robjohn Hello Sir. Can you tell me a good book to study theta functions and related concepts such as elliptic functions, elliptic integrals, dedekind eta functions etc. in depth?
 
2 hours later…
10:34
Hmm, I need help with this one: "show that if $\ sin x \neq 0$, then $\ \frac{1}{1 + cot^2 x} = sin^2 x $
Write out what $\cot$ is, in terms of $\sin$ and $\cos$.
Interesting, will try that
Hello could someone help me with this word problem: The capacity of a mobile telephone battery is the difference betweeen full charge and the charge at a given time x. the rate of charging is proportional to minus the capacity. Suppose that it takes one hour to go from 10% charge to half charge. How long does it take to achieve 90% charge?
10:53
Hi @AkivaWeinberger

« first day (2281 days earlier)      last day (3035 days later) »