« first day (3027 days earlier)      last day (1989 days later) » 

12:25 AM
I was eating): but i think i can catch your example
My apologize
 
 
6 hours later…
6:38 AM
A happening in conjuction with C.
Does it mean P(A/C) or P(A AND C) ?
 
7:15 AM
need more infomation, conjunction usually means "and"
[Random]
The following is inconsistent
Given $0+0=0$
then $$\lim_{n\to \infty} 0 + \underbrace{\cdots}_{n} + 0 = 1$$
Proof:
Suppose the above is consistent. Then using distributive algebra axioms, we have for each $n$:
$0 +\cdots + 0 = 0 (1 + \cdots + 1) = 0n$
However, zero product theorem said that $0n=0$
Suppose we disregard this and insists that:
$\lim_{n\to \infty} 0n = 1$
Then this means there exists some infinite element $x$ such that $0x=1$
However, assuming associativity holds, then we have:
$0n=0 \implies 0nx=0x \implies 0nx=1$
sorry typo
$0x=1 \implies 00x=01 \implies 00x=0 \implies 0x=0$
Thus contradiction
1
Q: Multiplying Infinite Cardinals (by Zero Specifically)

mathNotebookOn the Wikipedia page on Cardinal Numbers, Cardinal Arithmetic including multiplication is defined. For finite cardinals there is multiplication by zero, but for infinite cardinals only defines multiplication for nonzero cardinals. Is multiplication of an infinite cardinal by zero undefined? If s...

More poetically speaking, the above result told us that algebra transcends infinity
In particular we have the following theorem
Division by zero no-go theorem 1
Given any semigroup, the following holds:
$$01=0 \land 00=0 \implies 0\text{ has no multiplicative inverse}$$
 
7:35 AM
@Secret No-go is an interesting name for a theorem.
 
no-go theorems is not really a special name. These are common in physics when they impose limitation on what is or is not possible
E.g. No communication theorem in quantum mechanics told us that you cannot do superluminal signalling with entangled states
 
Zee
Except they aren’t really theorems in physics
 
Well eh... the no cloning and no communication theorem is really because of the hermitian properties of linear operators and tensor products, nothing physics really going on there
if you look at the proof, the whole thing is mathematical
(con't) I think the strongest of the division by zero no-go theorems I have proved is the following:
Aug 30 at 17:42, by Secret
Theorem: Given any division by zero magma $M_D$ where $0^2\neq 0$. Then it is maximally nonassociative (i.e. All possible associators terms will not be the identity map)
 
8:21 AM
@DanielML Let me know if you still need help
And next time please tell me when you stop being around
I kinda spent a significant part of my time helping you, I would have liked knowing you weren't going to answer for a while
 
@Secret please ping otherwise user is not notified.
 
ok
 
 
1 hour later…
Zee
9:30 AM
Math is cool man
 
Math is cool and math is hot, just like me.
 
why is the root of ab = -1 * root a * root b if a and b are negative in complex numbers?
i need a detailed explanation please
 
Zee
Cool is better than hot will
@TilakMaddy your questions is totally clear
Unclear*
 
10:01 AM
What to do if a question goes unanswered? No answers, no comments, just a tumbleweed badge!
3
 
10:33 AM
@JalajChaturvedi I starred your message because that happens to me too a lot.
 
11:20 AM
someone can take a look at my question: math.stackexchange.com/questions/3001155/… ?
@Astyx @mercio Hi !
 
Hi !
 
11:42 AM
hi
 
Complete means the span is dense right ?
 
12:03 PM
1
Q: If a function is continuous and injective on a closed interval is it always strictly monotonic

William Balthes(B)I note that if a function is continuous and injective, then very often the function is monotonic, and thus strictly monotonic, this is nearly always true? That is long as the domain and range are connected, and/or are real valued, closed and bound intervals, that this is correct, and that th...

Can you give an example of a function that is injective, but non continuous and of course not strictly monotonic. The point is to illustrate the importance of continuity
It would preferably be defined on an interval
I could find examples but only on unions of disjoint intervals
 
12:29 PM
@Astyx maybe that's an equivalent condition? i learned that complete= there is no orthonormal set that (properly) contains the complete set of orthonrmal vectors.
 
what's the complete set of orthonormal vectors ?
And what's an orthonormal set ?
 
orthogonal vectros with norm =1
$\{v_k\}$ is complete if it is orthonormal and there is no $\{u_k\} $ s.t $\{v_k\} \subset \{u_k\}$ (proper inclution)
 
Yeah I guess that's equivalent
 
@Astyx i think your defintion is equivalent though
later i will look at your answer, working on algebraic geometry questions right now, thanks! @Astyx
 
no problem
 
 
1 hour later…
1:37 PM
@AlessandroCodenotti are you still mad at AG ?
 
Nah
I'm doing an AG pset today
 
me too :P
@AlessandroCodenotti if i want to show that $S_d$ is contained in some homogeneous ideal $a$ , i think its enough to show that $x_i ^d$ is in $a$ , am i right?
 
What's $S_d$?
 
the homogeneous polynomials of degree d
are you studying from Hartshorne?
 
you also want $x_1^{d-1}x_2$
 
1:46 PM
The professor is following Mumford but I'm also looking at Hartshorne
 
ok so im doing chapter 1.2 right now
 
We're starting to talk about schemes which should be 2.2 or around that in Hartshorne
 
cool, my course is called "commutative algebra and AG" so we just started AG a week ago
so if i assume $\sqrt{a} = S_+$ i want to show that $S_d \subset a$ for some $d\gt 0$. so i know that $x_0,..,x_n \in \sqrt{a}$ so $x_i^m \in a$ for some $m$
now i want to show that $S_{m(n+1)} \subset a$
ah. got it, it is enough to check that $\Pi x_i ^{k_i} \in a$ where $\sum k_i = m(n+1)$ @LeakyNun
and that is easy.
 
2:29 PM
can anyone recommend me a good place to find the rules for learning / proofs in first order logic
 
i learned from "mathematical logic" by j.r shoenfield, dont know if it consider a good book but it was fine to me @user525966
 
I did try that book but found it really verbose (for me)
feels like he'd spend 10 sentences to talk about something when 1 would have been fine
 
2:46 PM

  Logic

This room is meant for discussion about logic, including found...
Experts there for good advice
 
3:01 PM
Let $\sum a_n$ be a convergent series over reals. If we remove infinitely many terms $a_n$, does new shrunk series convergent? (I know this is true for absolutely convergent series.) @LeakyNun
 
No
Take ${(-1)^n\over n}$
And remove the positive terms
In fact if the series is not absolutely convergent (i think that's called semiconvergent) then you can always remove an infinite number of terms and get a divergent series
 
oh! That's amazing. thank you very much. I knew that we may get any limit by rearranging semiconvergent series, but not this!
 
Well halfway through the proof of what you stated you can prove what I said
 
@Astyx So, removing all positive terms from any semiconvergent series gives divergent series, right?
 
yup
 
3:09 PM
ok, i am going back to revise proof.
hmm, in Rudin's proof, that is specifically stated.
 
I think you need to specifically state it
Like you prove the sum of positive terms diverges to $+\infty$
the sum of negative terms diverges to $-\infty$
And the terms go to 0
 
ok
 
Then you can go up by adding positive terms when you're under the limit you want to reach and add negative terms when you're above it
And because terms are getting smaller you'll stabilize at some point
 
yeah:) Idea of that proof is simple, but I had to check so many things to make sure proof is in fact correct.
 
If I recall correctly, there's some lemma which said that absolutely convergent sequences are those such that any subsequence is also convergent
 
3:20 PM
yeah that's true
 
3:33 PM
Consider $\Bbb R$ but for every rational number we add one copy, so for $q \in \Bbb Q$ we have $q^-$ and $q^+$
consider the order topology where $q^- < q^+$
are both copies of $\Bbb Q$ dense?
 
$Q^- \cap Q^+\neq \varnothing$?
 
$\Bbb Q^- \cap \Bbb Q^+ = \varnothing$
 
What does it mean exactly that a subset $D$ is dense in an ordered set $(O,<)$?
 
A generic definition of a dense set is that its closure is the whole set
so it will work for all topologies
So if $D$ is dense in $O$, then $\text{cl}(D)=O$ under that topology
Now I am guessing for any two elements in the same copy e.g. $\Bbb{Q}^+$ the usual linear ordering of the rationals has to hold...?
 
@AlessandroCodenotti For any $a,b\in O$ there exists a $c\in d$ such that $a<c<b$
 
3:44 PM
@AkivaWeinberger So if $O$ has two elements with nothing else in between it can't have any dense subsets? That's what was bugging me
 
Because that's what happening here, $q^-$ and $q^+$ have nothing in between if I understood correctly Leaky's order. It's like a subset of $2\cdot\Bbb R$ as order
 
How about, if we give $O$ the order topology, then $D$ is dense in $O$
and a basis the order topology is $(a,b)$, $(-\infty,b)$, and $(a,\infty)$
 
That should work for total orders, but I guess "dense" should make sense in posets as well
 
Hm
I'll be back but I'll think about it
 
3:48 PM
A dense set has to have any point in the parent set to be at least a limit point or a point in the set itself. That means any point in $O$ has to have some convergent net to it from $D$
But then what is $r ? s$ for $r,s \in \Bbb{Q}^+$?
 
Hey everyone
 
wait a sec...
20 mins ago, by Leaky Nun
consider the order topology where $q^- < q^+$
Order topology of the reals with the additional relation $q^- < q^+$?
hmm...
The basis of the order topology generates the same open sets as the open interval topology of the reals. So since there always exists some sequence in $\Bbb{Q}^+$ that will converge to its closure (which is going to be a copy called $\Bbb{R}^+$), and likewise for $\Bbb{Q}^-$, then both $\Bbb{Q}^-$ and $\Bbb{Q}^+$ should be dense in $\Bbb{R}$
hmm... where does the sequence $q_i^-$ where $q_i^-$ grows to indefinitely large in value, converge to. There is no smallest element $r$ such that $r < q^+$ for all $q^+$
 
4:20 PM
Being able to plot a representation of the rationals sure come in handy
know what
It is not clear if that can address leaky's question
 
I don't know. What do you mean?
@Secret
 
Sets of the form $S \Bbb{Q} + T \Bbb{R}$ are important in the study of dense linear orders, since every dense linear order is isomorphic to this
$$\text{Now we knew that Mats Granvik keeps yelling now we know for now we know who knows why reason now we know}$$
(multiply the above by $\Bbb{Q}$ to get the complete sentence)
to be exact, dedekind finitely many copies of it
 
@Secret I have to be completely honest with you. That was not me.
 
ok that's strange, cat typing maybe
 
4:39 PM
I didn't think I could get more confused but here we are
 
@Secret Is what you are writing about related to the Continuum Hypothesis somehow?
 
yeah, assuming GCH (thus I might have made an error with saying $T\Bbb{R}$)
actually any uncountable linear ordering in ZFC is a combination of 5 kinds of sets shown above
As for density I am not sure of the full case, but if the DLO set is countable, then it is isomorphic to $\Bbb{Q}$
 
4:55 PM
@Secret without endpoints*
 
right
not sure whether an analogous case holds for DLO with endpoints though (e.g. isomorphic to a disjoint union of half open intervals of rational points (i.e. sets of the form $[a,b) \cap \Bbb{Q}$))
 
5:09 PM
A countable dense linear ordering is isomorphic to $\eta$, $1+\eta$, $\eta+1$, or $1+\eta+1$, where $\eta$ is the order type of the rationals
 
5:21 PM
My apologies to Secret for being a bully. I thought it was funny at the moment and I laughed quite hard all by myself, but I regret it now.
@Secret
 
lol
 
5:47 PM
I really like staring at the moon
Maybe I should get a telescope
 
Hi DogAteMy.
 
Has anyone else noticed that the print in comment boxes on main gets scrunched together vertically? I can't figure out how to deal with this.
 
@TedShifrin hi
 
Hi Leaky.
 
6:00 PM
Can someone explain what is the condition for curves 'touching' one another?
 
@tatan solve them together and $\Delta = 0$. I mean take y from one equation and put that in other
Where $\Delta$ is discriminant
Hello @TedShifrin
 
A simple explanation would say that the slope of the tangents would be equal at that point but what about curves like x^3 and x^5 which have same slope at x=0 but they dont 'touch' at that point. So, what is happening?
 
oh you arent talking about conics.
 
Speaking strictly of chapter, this belongs to Tangents and Normal
 
its all on how you define touch.
if touch = meet only at a single point, then you need to solve the two equations together.
 
@LeakyNun I love bananas. I am also a banana.
4
 
I agree... but lets not make touch=intersect. Lets define 'touch' as the two curves lying locally on opposite sides of the tangent at the point of 'touch'
 
Can someone tell me what distance of point from line measured parallel to some plane means?
 
@WillHunting so you aren't a penguin anymore?
 
@tatan yes man. If you have like $f(x)$ and $g(x)$ then check for roots of $f(x)- g(x) = 0$. It should have only a single root so that they touch each other.
 
6:06 PM
@tatan assuming $a$ is not a critical point for any of the two curves, they are tangent at $a$ if their tangent vector at that point are colinear
 
@Astyx Am I not correct?
 
@LeakyNun I am King Banana Penguin living in Antarctica, jumping from iceberg to iceberg to change my timezone in a split second.
6
 
About what ?
 
About touching.
 
Oh
 
6:08 PM
1 min ago, by Abcd
@tatan yes man. If you have like $f(x)$ and $g(x)$ then check for roots of $f(x)- g(x) = 0$. It should have only a single root so that they touch each other.
Is this correct^?
 
Not in the way I understand his question
 
Hi@TedShifrin @LeakyNun
im trying to prove the triangle inequality using Cauchy-Schwarz
but im stuck at this point
 
@Abcd Can you explain how?
 
im gonna send an imgur to save typing all the latex if everybody is okay with that
 
@tatan If they are touching at one point only then ONLY ONE point $x_1$ satisfies $f(x_1) - g(x_1) = 0$ $\implies f(x_1) = g(x_1)$
5 mins ago, by Abcd
Can someone tell me what distance of point from line measured parallel to some plane means?
Can someone please answer this?
 
6:11 PM
@TedShifrin Hi Ted! What do you mean?
 
@Abcd nobody has to answer. Just hope somebody does.
 
when did I order
I just requested without even pinging anyone
 
@Abcd I never said you ordered
 
@Abcd You would expect a root of multiplicity $> 1$ though, if they not only intersect but are also tangent. This is because $f(x_1) = g(x_1)$ and $f'(x_1) = g'(x_1)$ in that case, so $(f - g)(x_1) = (f - g)'(x_1) = 0$, symptomatic of $f - g$ having a double root at $x_1$.
 
@JakeRose For $z\in \Bbb C$, $|Re(z)|\le|z|$
 
Anonymous
6:15 PM
@Abcd Does this help?
 
Anonymous
@TedShifrin Hi. :) I didn't notice anything as such. Do you have a screenshot?
 
@Blue Doubt in 0:50 . So PQ need not be perpendicular to the line?
 
Anonymous
@Abcd Right
 
@Blue Oh that was my main confusion since we generally measure distances using perpendiculars.
@Fargle right.
 
@Astyx. Hmm, but I don’t have a modulus around the Re part?
 
6:19 PM
You don't need it
Like, $x\le|x|$ for all reals
 
Ahhhhhh I see
the other issue I’m having is going from equality to inequality
im not quite sure how I ‘sub’ an inequality in
Actually
I think I get it
 
@Blue Let me try.
 
@Astyx Then latter part asks for conditions hinder which this holds
any hints?
U have they must be the same dimensions
but other than that I’m stumped
 
hinder ?
 
Hi all.
 
6:30 PM
*under
@CaptainAmerica16 hi
 
Blue, there :P
 
I learned implicit differentiation in class, I'm assuming concavity is coming soon.
 
Oh under which conditions this equality holds ..
 
@CaptainAmerica16 and how exactly are they related...
 
howdy @CaptainAmerica.
 
6:31 PM
Well in the series of innequalities you must have an equality instead
 
@JakeRose: You're talking about conditions on two vectors. Draw pictures.
 
@LeakyNun Idk, I just remember when I told Ted I was working on a derivative topic, and he asked me if my class had done concavity yet. I said no, but I assumed there was a reason why he asked, which is why I mentioned it now.
Today has just been super weird.
 
I don't remember why I asked about concavity. Maybe I was going to give you an interesting question to think about.
 
@TedShifrin I would definitely like to have something to focus on right now. Maybe I'll do Spivak for a bit.
 
I do have a cool question for you when you have learned about concavity.
 
6:34 PM
I'll be sure to remind you then :)
 
@Abcd Is the line contained in the plane, I hope?
The plane should be irrelevant, actually, but then you have a picture inside a plane, rather than inside 3-space.
 
@TedShifrin I am studying 3d geometry
 
Right, I figured. But the point and the line determine a plane, and you find the (perpendicular) distance in that plane, regardless. I don't understand what question they're asking you.
 
P(3,8,2)
$L \equiv \dfrac{x-1}{2} = \dfrac{y-3}{4} =\dfrac{z-2}{3} $
 
OK, so that's all the info we need. What are you talking about with the "measured parallel to some plane"?
 
6:40 PM
$\text{Plane} \equiv 3x+2y - 2z + 15 = 0$
@TedShifrin as blue's video showed, they want us to take a line parallel to the plane and then masure distance of P from L.
 
So the plane is not parallel to the line. I don't like the question at all.
The plane and $L$ intersect.
 
hmm
 
This makes zero sense to me.
 
@TedShifrin they do not want perpendicular distance it seems.
Like a slanted distance
of a point from a line.
 
In all my years of teaching such things, I've never seen such silliness.
So all I can think of is this: Move the plane parallel so that it contains $P$. Find the intersection $Q$ of that new plane with $L$. Find the distance from $P$ to $Q$. I have no idea why anyone would care.
 
6:49 PM
@TedShifrin How to find equation of projection of a line on plane?
 
Orthogonal projection, I assume?
 
they havent written anything like that in the problem.
Should I tell my thoughts?
 
$L \equiv \vec r = \vec a + \lambda \vec b$
Find point of intersection of line and plane
Call it $\vec p$
Desired line is:
$L_2 \equiv \vec r = \vec p + \mu b\cos \theta $
where $\theta$ is angle between plane and line
 
No, that doesn't make sense. You need a vector for the direction of $L_2$.
 
6:53 PM
@TedShifrin edited.
 
What you edited was not important. I understood what you were saying. But what I'm complaining about is still wrong.
 
oh you are complaining about $b \cos \theta$
 
Yes, you need a vector, not a number.
 
hmm, how to vectorise it?
 
You need to project $\vec b$ onto the plane. Hint: Think about projecting instead onto the normal vector to the plane.
 
6:54 PM
$\vec r = \vec p + \mu \vec b \cos \theta$
Does this work :P ?
 
Not at all.
It's still in the direction of $\vec b$.
 
Oh damn yes :/
 
Read what I typed.
hi @Tobias
 
@TedShifrin Hi
 
@TedShifrin didnt understand the hint
 
6:57 PM
Why not?
 
Oh wait
so we want $L_1 \times N$ ?
where N is normal
 
I don't like $L_1$ there. That's a line, not a vector.
No, that cross product isn't right.
Do you know how to project a vector onto another vector?
 
@TedShifrin hmm, i meant the parallel vector part of the line.
 
Direction vector, $\vec b$, is what you should have written.
 
How do I convert a two form $x dy \wedge dx + y^2 dx \wedge dz$ on $R^{3}$ to a vector field on $R^{3}$?
 
7:00 PM
@usukidoll: You should have been taught a formula for doing that.
 
which is?
 
I'm not telling you.
2
 
D:
 
@TedShifrin projection of a on b is: $(\vec a.\vec b) \vec b$ (not caring about magntiudes)
 
Almost right, but not quite ...
oh, but you have to care about magnitudes.
But OK.
So use that.
In this case, $\vec b$ is the vector you want to project, and I'm suggesting you project it onto $\vec N$. Then what?
 
7:03 PM
find perpendicular to that projection
but that would be so lengthy
 
What do you mean perpendicular?
You want the part of $\vec b$ lying in the plane. $\vec N$ is normal to the plane.
 
join the projection and the b vector
correct?
 
In other words, write $\vec b = \vec b_1 + \vec b_2$, where $\vec b_1$ is the projection on $\vec N$.
 
yes, thats what.
 
Hey everyone!
 
7:05 PM
wont be lengthy
 
Right.
hi Demonark
 
hi demonark
@TedShifrin thanks.
 
Sure, @abcd.
 
@Daminark Hi.
 
How's everything going?
 
7:10 PM
For who
 
trying to figure out something
 
For all of you
 
Oh...things are sort of ok, sort of not.
 
Ah, hope things improve soon
 
Yeah, me too. It's not that serious - more annoying.
@TedShifrin So I just realized - I don't think $a-b = -b$ is a true statement because $0-b = -b$
 
7:15 PM
Well, of course, @CaptainAmerica. It's true iff $a=0$.
 
Yeah, it was a silly question.
 
Yup, and you got perturbed at me for telling you to think. You need to get over that.
 
sigh I know.
 
 
1 hour later…
8:24 PM
This is from Pinter's abstract algebra book. I am not getting this part clearly: " *From
one point of view the set of the integers, with addition and
multiplication, forms a ring (that is, it satisfies the axioms stated
previously). From another point of view it is an ordered set, and
satisfies special axioms of ordering. On a different level, the
positive integers form the basis of “recursion theory,” which singles
out the particular way positive integers may be constructed,
beginning with 1 and adding 1 each time* ."
 
@UnknownMathMan which part are you uncertain about?
 
I am not getting how forming ring and satisfying ordering implies that - integer can be constructed with 1 and adding 1 each time .
 
they're talking about the positive integers in the last part
and there's no need to shout
 
I mean it is well known though..
 
they do not mean that those imply that.
 
8:31 PM
can you explain me please what they mean to say here?
@LeakyNun I am not a native speaker actually and don't know what 'recursion theory' is..
 
@UnknownMathMan Recursion theory is also known as computability theory and is a branch of mathematical logic.
@UnknownMathMan You do not need to know any recursion theory to study algebra. Just read that passage and move on, no need to worry about recursion theory.
 
@WillHunting Ok, I just wanted to know, how forming ring and satisfying ordering implies(as it is claimed to form the basis of 'recursion theory' and it is said that recursion theory implies) positive int. can be constructed with 1 and adding 1 each time.
 

« first day (3027 days earlier)      last day (1989 days later) »