« first day (2351 days earlier)      last day (2669 days later) » 

user228700
2:00 AM
Huh, although, if I take $T$ to be any non-zero number, then which $T$ would come out to be the fundamental period of the function? OK, I realise that this depends on the function but still...does anybody have any clue about all this?
 
It would be the smallest positive one for which $f(x+T)=f(x)$
Even if $f$ is nonconstant, if $f$ is not continuous, it's not necessarily true that there is a smallest such $T$.
 
user228700
Oh, OK. But the definition of a periodic function says that the given function should have a $T>0$ in its domain such that $f(x+T)=f(x)$ for all $x$ in the domain of the function. Is this correct or can $T$ be any non-zero number? (I understand that the fundamental period is always the least positive number but what about the definition of what a periodic function really is?)
 
You're asking whether $T$ needs to be positive?
 
user228700
Yeah.
 
The conditions are equivalent, since if $T$ is a period, so is $-T$.
Which means that if $T$ is a period, so is $|T|$.
So if there's any nonzero period, there's also a positive one.
 
user228700
2:06 AM
OK. It's just that the fundamental period is the smallest positive one, yeah?
 
if there is one
 
In fact, if $T$ is a period, so is $kT$ for all integers $k\ne0$.
 
user228700
OK, thanks very much ^_^
 
The above fact came from taking $k=-1$.
 
2:07 AM
keep on asking good questions Kaumudi!
 
user228700
@AkivaWeinberger Yes, I understand, thanks!
 
user228700
@GFauxPas :-P OK, thanks for the encouragement :-)
 
Youre welcome!
 
i think i'm going to have to come back to this proof later.
i'm making too many dumb mistakes.
 
what are you trying to prove Heather
 
2:52 AM
@heather: You know what to do ... But you're not on the top of your arithmetic/algebra game tonight.
 
Hi @TedShifrin How are you sir ?
 
Kasmir, you awake already?
 
Yes sir ! :D
could not sleep more than 7 hours
I usually have hard time sleeping when i have things to do =p
but woke up not tired so thats a good start
 
Good.
 
Ehm ._.i know you helped me alot but i still got few questions about the theory and am embarrsed to ask again =p
 
3:05 AM
Go on ...
 
well , let F = (P,Q ) be a continous vector field defined in an open simply connected region "omega" in the plane
show that if F has a potential U so the line integral of any curve Y in omega
is f (b) - f (a )
where b end point and "a" is start point
 
That's good. It's the FTC together with defn of line integral.
 
i know how to use all these formulas now , but the proof we did not do on class in a good way so it did not stick in my mind
why the need of simply connected region ?
 
You don't need simply connected. That's wrong.
 
arxiv.org/abs/1701.02084 this is kinda surprising
 
3:09 AM
Just use $F=\nabla f$.
 
hmm am sure i translated right
f here is the potential
we have a counter example on our book
 
Right. To Assume curl = 0 and deduce there is a potential takes simply connected.
 
when the criteria of del Q / del x = del P / del y , is not enough to garantee the existence of potential
 
The counterex is curl = 0 but not conservative.
Right.
 
exactly
if we look at this example
E = (x,y) / (x^2+y^2 )
 
3:13 AM
No.
 
we can find U = 1/2 ln (x^2+y^2) + C as our potential
wait wait wrong one
sorry
 
That's not the weird counterexample.
 
B = (-y , x) / x^2+y^2
this one :)
 
Right!
 
3:14 AM
If we have a finite field of characteristic $2$ must everything be a square
 
Yup @DogAteMy
 
any closed curve that encloses the origin gives us 2pi
but if it were conservative it should be 0
 
So I see that the set of squares is closed under addition (since $(x+y)^2=x^2+y^2$ in characteristic $2$) and multiplication
but I don't see why it would be the entire field.
 
If it winds around once, @Kasmir?
 
winds around ?
ehm yes
because we have singular point
we need to do another curve around that singularity
makes the result 2pi
 
3:16 AM
What's the kernel of the square map, DogAteMy?
 
Kernel meaning things mapping to $1$ or to $0$?
 
But if it goes around 3 times, Kasmir, you get $6\pi$.
 
$x^2=1$ implies $(x+1)(x+1)=0$ which implies $x=1$... I see.
 
yes but the point is that its not = 0
 
You Decide, DogAteMy :)
 
3:17 AM
so not conservative by definiton =p
 
And $x^2=y^2$ implies $(xy^{-1})^2=1$
which means $x=y$ by what I just showed.
I see.
So the only examples of fields of characteristic $2$ with non-squares must be infinite.
 
is that group theory ?
 
Or field theory :)
 
oh =p
am gonna take that course next week ><
group theory
 
I saw $\Bbb Z_2(x)$, the rational functions on zee mod two zee, as the counterexample.
 
3:19 AM
I thought you were an engineering major.
 
Since $x$ is not a square.
 
$(x)$
 
me ? no am math student second year
 
@TedShifrin Yes. Right.
 
Oh, cool, Kasmir. Good for you.
 
3:20 AM
:D
 
Then understanding the proofs is important.
 
yes !
 
Fun fact: $\pi$ is the root of a polynomial with integer coefficients...
...since $0\in\Bbb Z[x]$.
 
Oh?
Um, no.
 
Everything's a root of $0$ !
 
3:23 AM
You get a smack for that!
 
@TedShifrin if i get this question , like explain what a parametrization of surface is ? explain how to find the normal to that surface ? explain why the formula for area of surface look the way it is ?
what are the important points to write
 
It's the cross product of the partial derivatives again!
 
yes i know that but ermm
why is the maginitute of that
 
Back to area of parallelograms again
 
and the projection of that surface on xy plane
gives area
 
3:25 AM
No no
 
hmm
i knew i missed something here
from what we did was , from st plane to xy plane
 
But for a surface you're in 3-space ...
 
and the correction factor was magnetude of cross product of partials
yes (s, t, f(s,t) )
2 variables
each curvy area becames "plane" in xy
 
But that's not the general parametrization — what about a cylinder or sphere or ... ?
 
oh yeah ._.
thats not the same
let me think -.-
 
3:28 AM
Is anybody around who could explain the $\pm$ part of Steve's answer to math.stackexchange.com/questions/2089876/… to me? He keeps showing up on MSE but refuses to respond to my comments asking about it.
 
@AkivaWeinberger Oh, I forgot to say what I'm working on right now.
 
If you're not willing to answer follow up questions, then don't post answers.
They don't do anybody any good except your own ego.
 
@JessyCat It could be he hasn't been online in the past two hours
 
The simplest version of the problem goes like this. Consider the following first-order system of ODEs: $$i\frac{d}{dt}\begin{bmatrix}\psi_1 \\ \psi_2 \end{bmatrix}=\begin{bmatrix}-\alpha t & 1 \\ 1 & \alpha t \end{bmatrix}\begin{bmatrix}\psi_1\\ \psi_2\end{bmatrix}$$
 
@AkivaWeinberger he was just online 1 minute ago.
 
3:34 AM
Never mind, then.
 
You can tell how long ago somebody was online by looking at their profile.
 
The $\pm$ is surely because the center is $\pm I$.
 
If I take the boundary condition to be $\Psi=(1,0)^T$ at $t=-\infty$, then this is a boundary value problem with a definite solution.
In particular, I can ask for the behavior at $t=+\infty$.
It turns out to be equivalent to a 'parabolic cylinder function' (a certain special function) with known asymptotics, so one can look those up.
And they turn out to be simpler than one would expect.
The questions are 1) how to do this calculation in a non-horrible way, 2) how to generalize it to more complicated-looking operators that are still of this same sort.
 
They should have spring hats too. It's always a good time for hats.
Like $\hat{y}$
Or $\hat{Jessy}$
It's a Jessy Hat
 
3:54 AM
$\rm\widehat{Jessy}$
2
 
Bad joke: What's a nice present for a math friend?
A $\hat{\nu}$.
 
There we go.
@JessyCat A (Jessy) Cat. ^
 
4:11 AM
Yippee!
or should I say meow
 
Hm. Does the chat allow GIFs?
Oh, wow.
 
Heavy duty
 
neat.
 
A nu hat. Bwahahahaha
A $\hat{\nu}$ for a $\mu$-cat
Anyway, I'm going to get some zzzzzzz
G'nite
 
4:49 AM
@Semiclassical Hurray for nu math / it won't do you a bit of good to revu math / ...
 
Last night dream: Pseudoabsorbers:
$\exists a, axa=a^2, \forall x$
It is easy to see that absorber implies pseudoabsorbers, but what will happen if a is not an absorber, no idea
Inspiration pathway: Probably from too much meadow reading
 
@AkivaW What's new today
 
5:24 AM
This is the season of illness and I hate every second of it, every third, forth, fifth, and so on.
To everyone here, I wish you all well.
 
why is this season of illness
 
5:39 AM
Because people get sick during the winter.
At least, where I am. Some places in the world are having a fun summer right now, I bet.
There are also heavy implications in my posts that I'm a victim of the season of illness.
 
5:53 AM
@akiva you can generalize your squares in characteristic $2$ question easily to finite fields in general if you know that their multiplicative group is cyclic and how to count squares in Z/mZ
 
6:03 AM
Hello all! :D
 
6:59 AM
in The h Bar, 5 hours ago, by Kaumudi. H
I have a very quick question about the definition of periodic functions. My textbook says that the period of any function, $T$ must always be greater than zero while other sources that I have found say that it is sufficient that $T \ne 0$. The latter seems to be correct but I wanted to make sure; so which is it?
Inspired a weird idea:
$f(x)=f(x+T)=-f(T+x)$
noncommutative periodic functions
 
what means "note that"? i mean, what would be a synonymous meaning?
because this sounds cringy in other languages
 
"Note that" $\to $ "Be aware that"
@Null Forgot to reply to you.
 
@Axoren where did you want to reply to me?
 
@Null "Note that" $\to$ "Be aware that"
That's all. You asked a question and I answered
 
@Axoren ah ok, I thought you wanted to answer another question hehe. Thanks.
 
7:05 AM
Without using the @, it's likely you wouldn't have seen it. No prob.
 
[Abstract algebra] One of the useful low level ways to test how a new axiom would behave is to assume the structure is a field and then plug in that extra axiom, and then track all the consequences (contradictions or not) that happened. Then modify the field axioms until no contradictions arises. Then start wondering what that new axiom means geometrically
 
@Secret Could you give a non-trivial example?
Like, a simple one?
I guess that's asking too much, now that I think about it. A non-trivial yet simple example.
 
I am not sure, some of my attempts so far had gave me trivial rings.
However I can illustrate how it works. For example consider the ring of integers $\mathbb{Z}$ and let's modify the distributive law so that + distribute over *, and left anything else unchanged. Then one can deduce that every integer will become idempotent under multiplication. Howeve for this example, eventually you get $1x=1$ thus the whole thing just collapse into the trivla ring. I have not explored further what axioms I can relax to avoid the collapse, though, because it is not my main focus atm
Well let's see, suppose in the field of reals we want to make + to self distribute. i.e. a+(b+c)=a+b+a+c. Then it is easy to see that 1+0=1+(0+0)=1+0+1+0=1+1=2
one immediate consequence is that associativity in + has to be broken in order to avoid contradictions
 
What if we try to construct $(\mathbb Z, +, \oplus)$ as if it were $(\mathbb Z, +, \times)$ and $\oplus$ has the same identity element as $+$?
It can't be a field because it lacks the distributive law, but would it still have some geometric meaning?
 
7:22 AM
Let's see:
So basically we have $0=1$ thus for simplicity we just call our identity 1
Now surveying our usual axioms in the ring of $\mathbb{Z}$, modified with our suggested axiom

1+1=1
1*1=1
a+1=a
a1=a
(ab)c=a(bc)
(a+b)+c=a+(b+c)
ab=ba
a+b=b+a
a(b+c)=ab+ac
a-a=1
ab=1 => a=1 or b=1

Begin analysis:
First off, we lost the induction property of $\mathbb{Z}$ as 1 is now idempotent in both + and *
 
We can get something similar back, couldn't we?
Inducting on the evens then the odds?
n+2, etc
Whatever that means now
I'm assuming we still have the following: $a \le 2$, $a < a+2$
now that we have $a = a + 1$
Are we still saying that $2 = 1+1$?
 
We cannot say 1+1=2, as in this new system 1+1=1, thus if you include induction then 2=1 and the structure collapses into the trivial ring
we can however somehow introduce 2, since a1=a
 
But then haven't we just shifted the role of $0$ to $1$?
Okay, hold on
 
no, your suggestion on 0=1 basically merged the roles of 0 and 1, which is why induction is gone. But because a1=a is still intact, 2 is no longer related to 1 and we can actually have some freedom to set e.g. expressions for 2+2 etc.
 
Yeah, we've got both of them functioning equivalently now
I'm imagining real space but everything kind of collapsing in the region of $[0, 1]^d$
 
7:32 AM
Also setting 0=1 actually does not seemed to compromise any of the original axioms except for induction at least for the element 1. 2,3 etc. now become somewhat independent of 1 thus there's actually a lot of freedom on how we want to set rules for these elements, and they will all be defined a1=a
so the next logical step to check whether something else will break is to plug a1 into all these axioms and see what happens
 
How would we even notate fractions in this system?
 
that will tell us more about the nonidentity element a in this modified integers

uh, fractions are a bit too far right now. While it is true that fractions are usually defined via an equivalence relation of (x,y)(z,w)=(xz+wy,zw), I have not gone far into fractions yet
 
It's consistent, unless I made some mistake.
substituting $a=a1$, I mean
 
@Secret Induction is a hell of an axiom to sacrifice
 
We might be able to somewhat recover it via a multiplicative version of it (that's my initial attempt in trying to explain to akiva what a reversed ring is), but it all depends on how $a$ behaves in this system. So far

a1+a1=a(1+1)=a1
a1+a1=(a+a)1=a+a
=>a+a=a

a1*a1=a1a1

Thus all a are idempotent under addition
so geometrically, everything becomes isolated at least under addition wrt itself
 
7:38 AM
Wait, what?
But $2+2\neq2$
 
We get 2+2=4 in the integers because of the induction axiom leading to 1+1+1+1 which we defined it to be the numeral 4. But here, induction is broken, thus it is not suprising that $a+a\neq a$ will fail
 
Can't we induce by twos starting from $2, 3$ and $-1, -2$?
$4=2+2$
$5=2+3$
$6=2+2+2$?
 
whatcha all up to
 
@SAWblade too delirious to know anymore. $0=1$, make it work.
 
easy. axiom #1 -> there exist two things called $0$ and $1$. they're the same
bam consistent system
i find that the more george michael one listens to the more successful at math you become
 
7:45 AM
@Secret I'm going to have to bail. I'm not in the right mind for this right now. It seems like there's still a way to get something that works enough like induction to replace that lost feature, but I'm not in proper sorts to wrestle with it.
 
a1*a1=a1a1=aa1=aa
=>aa seemed to be quite independent

a1(a1+a1)=aa+aa=(a+a)a=aa
=>a(a+a)=aa

a-a=1=(1-1)a=(1+1)a=a
=>contradiction, hence no -a

aa=1=>a=1
rest seemed fine

---
Possibly,, since 2 is not dependent on 1. Thus we might have at least two generators for this structure <1,2>
Yeah, they are not easy, it takes time to think and experiment (that's why it is low level method)
 
secret is there a reason you're not using TeX :0
 
uh, the conversation flows fast and the expressions are not too complicated that they are still readable with mathjax?
 
fair enough
just curious, i rarely ever see someone type math without it here
 
I do use mathjax in most of my messages, though
 
7:48 AM
$likewise$
so what're you up to tonight then
 
It's 18:49 in my place, thus it is not very late. I should be continuing on reorganising my proofs on some zero term algebras (algebras that involve division by zero)
except some of the proofs never look clean enough
 
interesting, you're 8 hours behind me :0
and i feel you
some proofs just always feel nasty
i'm supposed to be either sleeping or working on research
but alas here i am
 
Being trained in the natural science and the scientific method, I have a tendency to try to poke holes at my own proofs, until it is perfect
so basically, I don't really think like a normal mathematian would
 
excellent, that's an edge
 
I also not just satisfied at finding a counterexample, I will try (if it is feasible) to find the whole set of counterexamples, thus finding counterexamples are really finding a set of examples obeying a proposition that is required for a given counterexample
That is, I like to understand what makes a pathological thing pathlogical, what makes an unatural structure unatural and so on
 
7:55 AM
ah, i see
not just seeing that something is incorrect but trying to understand exctly why its incorrect
good form
so how does one divide by zero
mm, actually i should sleep, i have no reason to be up at 3:00 a.m.
i wish you luck on further perfecting your proofs
g'night
 
If you are talking about a multiplicative inverse of zero, that's a work in progress and no one had done before as far the literature goes.
If you are talking about the broader scope of defining n/0, then the best known literature examples are meadows and wheels

g'night
basically keeping 0+0=0 and allowing zero inverses can result in a lot of elements becoming additive identities. Still trying to tame them so that they don't give me null semigroup liek structures
 
 
1 hour later…
9:22 AM
I'm having some problems figuring this one out, I'm pretty terrible with probability.
For a standard deck of cards, what is the probability of taking two cards of the same suit?
I thought this probability would be the same as P(A | A) where A is taking a card of some arbitrary suit
 
@OverlyExcessive Once you have taken one card, how many cards are left, and how many of those have the same suit as the one you took?
 
12 / 51
 
To me it seems like it should be (13 / 52) * (12 / 51) but I don't seem to arrive at the right answer
 
that is the probability of drawing two cards of a specified suit
 
9:26 AM
Right
The question is, if I take two cards from this deck what is the probability that they are of the same suit. Now wouldn't that be the same thing?
 
I don't understand the distinction. Can I not solve this problem through conditional probabilities alone?
The first card I draw has to be of some suit -- now it follows that the probability of drawing another card of that same suit is 1/17.
 
where did 17 come from?
 
Probability of drawing a card of Hearts for example, 13/52, probability of drawing another card of suit Hearts, 12/51, (13/52) * (12/51) = 156 / 2652 = 1/17
Ok the problem states that I'm drawing 2 cards but I don't think it would matter if I draw two cards simultaneously rather than one by one
 
right, but as you just noted, you found the probability of drawing two hearts, not the probability of drawing two of the same suit
think of i t this way: the first draw cannot "fail"
 
9:42 AM
Yeah okay
I think I see what you mean
It seems to me that the probability would be quite a bit higher than 1/17 then.
 
it might also help to think of a smaller example: flip two coins and consider the different outcome spaces of "same result" versus "two heads"
yeah, I don't see where you got the 1/17 from
 
But outside of computing the cartesian product for two cards in a 52-card deck I don't see how I would calculate this
 
ohh, right, 1/17 is the probability of drawing two cards in hearts
by noting precisely that it does not matter whether you draw one at a time or draw them at the same time
so you can draw one at a time and only use the first draw to determine which cards will be a "hit" for the second draw
(also worth knowing: if you have a collection of things with an equal number of each thing, then the probability of getting two of the same thing is very close to 1 in the number of different things)
 
but then wouldn't the probability just be 12 / 51 ?
oh it was
Wow, thanks :)
 
right (which is close to 1/4 and there are 4 different suits, so the intuitive idea checks out)
the "close to" I mentioned can be made precise, but that is not so important
 
9:48 AM
Could I have figured it out algebraically? I mean seeing as it is basically the conditional probability without the product of the first event
 
what would the condition be?
yes, but you would need to make it conditional in the right way
 
Well I mean isn't it just P(B | A) - P(A) ?
 
where A and B are what?
 
Where A is the probability of drawing a card of some suit, let's say hearts
And B is the probability of drawing a card of that same suit
 
Hrmf. I need to stop trying to prove things at 1 am
 
9:55 AM
$p$ is a prime number and we know all of its primitive roots, id like know how i can find primitive roots of $p^2$, $p^3$, ..., $p^k$?
 
@OverlyExcessive that makes no sense to me why it would give the desired probability
 
Haha I probably got it wrong then :P
 
@euclid Hmm, as far as I recall, if $a$ is a primitive root mod $p$ then either $a$ or $a+p$ will be one mod $p^2$ (but that might be me misremembering)
 
I was just thinking that since you are not really depending on the first event you could just take that out of the equation
 
@TobiasKildetoft you are right but you know $p^2$ has $\phi(p(p-1))$ primitive roots so there primitive roots except $a$ and $a+p$.
 
10:02 AM
@euclid Right, but once you know one primitive root, you know them all (modulo factoring numbers)
 
@TobiasKildetoft explain please
 
@euclid If $a$ is a primitive root, then the other primitive roots are precisely those of the form $a^m$ with $m$ coprime to the order of $a$.
 
@TobiasKildetoft yes, and by this i can find other primitive roots of $p^2$?
 
@euclid Once you have one, yes
Just received the most lazy scam mail ever. I thought everyone at least tried to spell the name of the company they impersonate correctly now, and at least tried to hide the fact that the email was not sent from that company and that the link did not go there
 
@TobiasKildetoft thanks,come back to beggining, is there any way to being sure that $a$ is primitive root or $a+p$? or only we have to test it?
 
10:10 AM
@euclid That part I don't know unfortunately
s23.postimg.org/l9vscyuob/… here is a screenshot of it. It is a thing of beauty (and even if my Swedish is not great, even I was able to recognize that this was really garbled)
 
@TobiasKildetoft thank you very much anyway. it is same for other numbers like $2p^2$,...,$2p^k$?
 
not sure
 
do you know any reference for them.i am self-studying and i dont have i dont know good book for them.
@TobiasKildetoft
 
@euclid No, sorry
 
@user91500 hi,i am beginner in analytic number theory.i have some questions.i hope you help me
@user91500 it is about primitive roots
 
10:35 AM
@Euclid Just ask your question, then anyone here that have enough time probably can help you.
 
i want to know how we can find primitive roots of $2p^2$,$2p^3$,..,$2p^k$ when we have primitve roots of $p$,$p$ is prime
 
10:47 AM
@TobiasKildetoft can you explain the first paragraph of this answer? math.stackexchange.com/questions/1860079/…
 
which one? You linked the question
 
yes,first answer
 
Not sure what there is to explain. The claim is that one side is 0 and the other is 1 in case $p$ divides $n$
 
i meant first paragraph
 
@TobiasKildetoft I have Micorosof Windows 7 as my laptop's OS.
 
10:52 AM
fix n

and define

f(d)=∑o(χ)=dχ(n).

Let's show f
is multiplicative. First off, let g be a generator for (Z/pZ)× and write n=gk, then let ψ be a generator for the group {χ:χd=1}, in which case we may say o(χ)=d⟺χ=ψe for a unit e mod d. As ψ(g) is a primitive dth root of unity, the values ψ(g)e (as e ranges over units mod d) will be all primitive dth roots of unity, i.e. all ζ with o(ζ)=d

. Thus

f(d)=∑(m,d)=1ψm(gk)=∑o(ζ)=dζk.
 
@BalarkaSen Yeah, probably the name Microsoft in the signature is a typo :)
 
I still felt like holes can be poked on this proof:
Proposition: Given a semigroup $S$. If $\exists a,b \in S, ab=1$ and $\forall x\in S, 1x=x$ and $bx\neq by, \forall x,y \in S$, then $a: x \to ax$ is bijective and hence a permutation.

Proof: Since $bx\neq by, \forall x,y$, $\exists z \in S, bx=z, \forall x$ (i.e. $b: x\to bx$ is injective. Next multiply by $a$ on the left to get $a(bx)=az$. By associativity $a(bx)=(ab)x$ and by $ab=1$, $abx=1x$ and then by $1x=x,\forall x$, $1x=x$. Since $1$ is a left identity, it is injective wrt all $x$. Hence $a(bx)=1x=x$. Since $1$ is injective and $bx\in S$ is unique due to injectivity of $b$, $a$ is
Uh, wait a sec...
> A bijection from the set X to the set Y has an inverse function from Y to X. If X and Y are finite sets, then the existence of a bijection means they have the same number of elements. For infinite sets the picture is more complicated, leading to the concept of cardinal number, a way to distinguish the various sizes of infinite sets.
 

« first day (2351 days earlier)      last day (2669 days later) »