« first day (3498 days earlier)      last day (1818 days later) » 

16:21
@TedShifrin oof. I think my immediate reaction is the same as yours, tho: get rid of $y(t)$ in favor of $x(t)$ and solve for $t$
But boy does x(t) seem horrid
16:36
Consider a set of fuzzy (partially true) truth values $T=\{t_0,t_1,t_2,...,t_n\}$ with a partial map $\psi:T\to J.$ Where $0\le t_0,t_1,...\le1.$ How do you describe the notion of the target space having a phase space of possible values?
for example, an element $t_0 \mapsto j_n$
$n=1,2,3$
at some time $q=q_0.$
$\int_0^\infty e^{n!(x^{1/n})}dx$ seems like a tricky integral
but then at time $q=q_1$ the target space could be different.
($n\in\Bbb N$ for the above, if that needs clarification)
yeah it does seem tricky
For all the values of $n$ I've tested, it comes out to 1, actually
But it's been a while since I've done any real calculus so all my testing has been with a computer, and it feels like this'll be tricky (it might not be, of course)
16:53
I think the substitution $u=n! x^{1/n}$ sorts it out
Ah, yeah, that'll make it easy
Reducing it to computing an integral of the form $\int_0^\infty e^u u^{n-1}\,du$, which is just $n!$
That said, I don’t see it canceling down to 1
Except in the cases n=1,2
Oh, wait. Is that n! or (n-1)!
you sound excited
Lol
Actually, it’s neither: that integral blows up :)
Ah, wait, forgot the negative
16:59
Yeah, I figured
I was just crunching the numbers again and was getting it blowing up, too
Anyways, the integral I quoted should indeed be (n-1)!
Yeah, plugging in $\int_0^\infty e^{-(5040x)^{1/7}}dx$ gives 1
Wait, another typo
$\int_0^\infty e^{-(n!x)^{1/n}}dx$
Raise the $n!$ to the $1/n$ power as well
17:02
Yeah
it's probably a Gaussian integral
No
It’s just the integral representation of the factorial
Specifically, it’s what you get if you start with $n!=n\int_0^\infty e^{-u}u^{n-1}\,du$ and substitute $x=u^n/n!$
(And then divide both sides by n!)
I kinda hate integrals like this tbh: start with something easy and use a weird substitution to make it hard to recognize
Heh, I just stumbled upon this while fooling around with wolframalpha
Specifically, I was trying to take an exponential function over the domain $[0,\infty ]$ and trying to normalize it
17:21
Are there any applications for, a convolution on the the analytic structure of two (or more) transversal manifolds?
convolution, I know is a nonlinear operation
in this context (manifolds) it is a generalization of the warped product
In mathematics and physics, in particular differential geometry and general relativity, a warped geometry is a Riemannian or Lorentzian manifold whose metric tensor can be written in form d s 2 = g a b ( y ) d y a d y b + f (...
I guess I sort of found an application.
0
Q: convolution of Riemannian manifolds

geocalc33I would like to understand the first section of a paper titled: "More on Convolution of Riemannian Manifolds," by Bang-Yen Chen. Michigan State University. http://emis.impa.br/EMIS/journals/BAG/vol.44/no.1/b44h1chn.pdf (first two pages). The notion of convolution products is defined as follows:...

In case you would like to read^
18:02
hi folks, could you tell me what is 0.954 in 1 significant figure?
Is there a consistent extension of Morse-Kelley set theory that is finitely axiomatizable again?
18:25
hey all, once again. I've started realizing I probably won't go back to university (given doggo, full time job, and living in slovakia), but also that I have a lot of free time I'm looking to fill up with things I find fun. So I thought it could be interesting to go over the lecture notes and exams of the various lectures from MIT, in order (disgarding outdated ones): ocw.mit.edu/courses/mathematics
is this a crazy thought? I'm a developer for a living, and I'm also trying to minimize screen time because I get too much of it from work, and my eyes can start hurting
would I be better off buying a text book and going through it?
@towc the MIT courses are nice self-contained courses that tell you what to learn, so this is an advantage I guess
buying a textbook and just trying to read it cover-to-cover often results in failure
I have a couple textbook pdfs (mostly for development topics) on my phone, but I never finish them
I really dislike having to read from a phone, and I'd guess it's the main reason I don't keep going
Right, I think textbooks are good to be used in conjunction with a course
but I'd also not enjoy carrying a book around
and if you know how to use them properly then they are good by themselves, but that's a skill in and of itself
18:37
in highschool I couldn't focus on textbooks at all. I did 1 year of uni, on computer science, and all we needed were lectures, which are a format that works for me. But I don't like the idea of spending my free time watching videos
so I'm not sure what alternatives could fit my needs
A combination of all of the resources in such an MIT course is good
ideally I'd find a textbook that makes you think most of the time, and after you've thought, the book tells you how a problem is conventionally approached, poses another problem, and repeats the process
this should be able to fit on my phone, with me spending most of my time scribbling notes on paper and feeling productive, right?
can you think of anything like that?
I guess the "Schaum's Outline" books have that sort of feel
They have a lot of worked problems and a lot of "Supplementary Exercises"
I doubt something like this would exist that would cover as much ground as the MIT courses, but it's a start
Yeah the Schaum's outline books are good as a supplement to a course but you should always consult the texts that are quoted/used in the course
@towc right, I used this in calc courses in my undergrad
loads and loads of problems
it does go a bit slow
I think I did all of this in highschool: mechanical.uonbi.ac.ke/sites/default/files/cae/engineering/…
but I've definitely forgotten most of it :)
You can just pick problems out of it, they're not really meant to be used as a course tex
t
any explanation why
this is my attempt
https://math.stackexchange.com/questions/3561562/how-to-compute-the-average-power-for-this-signal/3561582#3561582
well, $-\frac{1}{\omega_0}\le \frac{\sin(2 \omega_0 T) \cos(2 \theta)}{\omega_0} \le \frac{1}{\omega_0}$ does the job
18:53
I wonder if I can get a kindle or other non-harmful device to render it
Sanity check: $k[X,Y]$, $k$ a field, is noetherian and factorial, but e.g. $k[X,Y]/(Y)$ is an integral domain and not a field, so $(Y)$ is a non-maximal prime ideal, i.e. $k[X,Y]$ is not a Dedekind domain.
ja stimmt
Dedekind domains, as I recall, intuitively correspond to coordinate rings of curves ... so of course the coordinate ring of the affine plane will not be one.
@Thorgott but in the answer, he assumes 1/2T exist! why he distributed the limit?
As far as I know, the limit properties are used for limits that are exist.
well, the answer is very sloppy
the inequality still does the job though
you do want to distribute the limits, so you have to check the individual limits exist
@CroCo: This is the same sort of thing you and I spent an hour on a week ago. It's the squeeze principle to establish the second limit.
19:08
@TedShifrin I got a teaching interview (7th/8th) grade math/science
I'm not asking for $-\frac{1}{\omega_0}\le \frac{\sin(2 \omega_0 T) \cos(2 \theta)}{\omega_0} \le \frac{1}{\omega_0}$ in isolation. That is the problem. I don't know how he jumps to the answer
@TedShifrin you are right but this is different problem.
Interesting, @geocalc.
@CroCo: It's the same phenomenon, different problem.
Why is it interesting? lol @TedShifrin
I'm trying to solve several problems in this topic
Hey @TedShifrin, I don't know this perspective but I expect I'll learn it in AlgGeo1 :D
thanks
19:09
@TedShifrin could you please shed some light on it if your time permits
0
A: How to compute the average power for this signal

ArueraThe point is that $$ -\frac{1}{\omega_0}\le \frac{\sin(2 \omega_0 T) \cos(2 \theta)}{\omega_0} \le \frac{1}{\omega_0}, $$ and this is true for all values of $T$. Therefore $$ \lim_{T\rightarrow \infty} 2 T+\frac{\sin(2 \omega_0 T) \cos(2 \theta)}{\omega_0}=\lim_{T\rightarrow \infty} 2 T $$ and $...

as I said, distribute and see if the individual limits exist (spoiler: they do)
that answer is not good
The answer left out the squeeze principle step.
But it's clear that the answerer was applying it. If $a\le f(t)\le b$, then $\frac aT\le \frac{f(t)}T\le\frac bT$, so as $T\to\infty$, since $a/T$ and $b/T$ go to $0$, so does $f(t)/T$.
well, it also asks to do $0\times\infty$ by the end
I just wrote a comment saying it doesn't make sense.
There are lots of crap answers on this site.
yeah
but the idea still leads to the correct answer
19:15
Someone posted an answer to the question I asked for help on yesterday. I started working through the posted answer and got stuck on the second line. I asked the answerer how he arrived at it, and ultimately he realized it was wrong and removed the answer.
@Edward Dedekind domains have Krull dimension 1, so they are one dimensional affine schemes so it makes sense that they should be curves or something close even though I don't know the details of this story
@CroCo: I will post a correct solution.
yeah, I only know the purely "algebraic" perspective, since I've not done any geometry llol
My days of knowing the algebra go back 45 years.
19:18
Isn't there a section on this at the end of Neukirch chapter 1?
Yeah I think so
seems like I might want to invest in this sony e-reader: amazon.com/Sony-DPT-RP1-13-Digital-Paper/dp/B072DXXXN1
great, doesn't support linux, but there's some scripts that can help: github.com/janten/dpt-rp1-py
Is it possible for partially true information to yield completely true information after a certain amount of time in the context of mathematics
using a statement to equal a logical statement
I guess it's done via Godel
I want to show that locally, a system can obtain a classical and well-defined result that supports a well-defined set of logical statements
what I mean by locally, is that a system $A$ is a subsystem of $B$ and $B$ is global relative to $A.$
I'm thinking graph theory would be the best way to go about this
19:41
@CroCo: Two solutions posted.
19:57
@TedShifrin thank you so much.
The first is the right intuition, but still made rigorous with squeeze. You see that squeeze is showing up everywhere in your work!!
@TedShifrin true. I've read the entire chapter about limit in Calculus by Steward but hard and practical problems usually are not covered. Meanwhile, I'm reading some signals and systems and these problems pop up.
This is all elementary with the squeeze principle (theorem), which is certainly covered in the calculus books.
But notice that someone else posted an answer after mine, totally ignoring the ("rigorous") question of the difference between all $T$ and all integer multiples of the period. That's why I wrote that stuff carefully for you. But in practice signal engineers won't ever worry about that.
It seems the guy uses a theorem that says if a signal is periodic with fundamental period $T_o$, then the normalized average power P of the signal is the same as the average power of the signal over any interval of length $T_o$.
I like a rigorous approach like the squeeze theorem.
dear all
hello
I'm new
20:14
@lindaOiladali you are welcome
i have a function defined by $ f(x)=[x]+\sqrt{x-[x]}$ where [.] is the floor function, I proved that $f([k,k+1[)=[k,k+1[$ for all k in Z, and f is differentiable on R\Z, how to deduce the monotony of f ?
@CroCo thank you
20:26
do you have an idea
20:49
How rigorous a proof do you need? $\sqrt x$ is increasing on $[0,1]$ and the graph makes it clear that we shift up one unit when we go to the next integer, so that it continues to increase.
On any interval $[n,n+1]$, the fact that $\sqrt x$ is increasing does it. If $a<n<b$, then $f(a)<n<f(b)$ does it.
Hey @Paul long time no see
Hey @AlessandroCodenotti It has been a while
How are you? Still thinking about GGT?
Yup, still doing GGT and stuff related to mapping class groups. I am alright, tired of typing and went to see what was happening here. Doing set theory still?
Kinda, I ran out of set theory courses I can take and I'm not doing set theory for my thesis since the logic group was disbanded following the retirement of its only professor
Oh actually, I'm doing some stuff somewhat close to GGT for my thesis, about asymptotic dimension of metric spaces, so far I've been reading some work by Dranishnikov mostly
21:03
Oh, this is your masters still? Cool, I have looked a bit at asymptotic dimension, don't much. There is actually this space I thought about for a tiny bit that I think should have infinite asymptotic dimension but haven't really put the time to formulate an argument (in part because it would only be interesting if that ended up not being true)
Yes, I should finish my masters this summer term
Basically there's a weird compactification of proper metric spaces which is called the Higson compactification and Dranishnikov proved that for proper metric spaces of finite asymptotic dimension the asymptotic dimension of the space coincides with the Lebesgue covering dimension of its reminder in this compactification
My thesis advisor used this result to prove that if you quotient a proper metric space by a finite group acting by isometries the space and the quotient have the same asymptotic dimension, but he was using Dranishnikov's result as a blackbox so he asked me to learn how that works
If you are interested it is the ray graph for the plane minus a cantor set: vertices are simple rays from and to infinity up to isotopy, edges between if there are no intersections between them(up to isotopy).
uh weird
I never worked with isotopies
Eh, you don't really have to think about it too much, just you are allowed to move the ray around in the obvious topological ways (or you can think of homotopy) as long as it goes to infinity (or you can think of it as a sphere minus an isolated point and a cantor set and the rays go to the isolated point)
I don't think I understand the without intersections part, in the sphere picture can't I make any two such rays cross close to the isolated point? The missing cantor set is contained in a compact subspace so it won't be an issue for those homotopies
21:17
oh, forgot the definition (there is a related graph like the one), one of the points starts at infinity the other end of the ray ends on the cantor set
Oh ok, I see how that works now
This looks like a huge graph
Oh yah, there are similar graph even just on closed surfaces (the curve graph), which are locally infinite and stuff like that. Somewhat surprisingly they end up being unbounded (including the above ray graph) and having interesting geometric properties like being gromov hyperbolic
The finite type ones have finite asymptotic dimension
I see, interesting
One of the authors Greg Bell has done a few papers with Dranishnikov
Actually Dranishnikov was his advisor
@PaulPlummer Oh, definitely, I keep seeing his name
@PaulPlummer I didn't know but that makes a lot of sense
21:29
Are you going to be doing an expository on the Higson compactification result?
That's going to be part of it for sure, I already typed up half of the result (Dranishnikov proved the two inequalities separately a few years apart from each other)
But maybe I'll also have something more to put in the thesis, we'll see
Do you have a phd program lined up if you are graduating this summer?
Still working on that
But I definitely want to do set theory for my phd
Sweet, know what kind of set theory you are interested in?
21:46
$F$ and $Q$ are fields. Does duality of $F$ and $Q$ imply well-definedness of $Q?$
If a field is dual to another field, the solution of the field can be obtained from the dual field by the duality transformation.
@PaulPlummer set theoretic topology mostly
Also infinitary combinatorics in the forms of cardinal characteristic of the continuum and problems concerning the tree property is very interesting
But to be honest all set theory I've seen looks quite interesting, apart from maybe inner model theory
@geocalc33 let's choose one problem from section 1, I'm finding those most difficult. Then in section 2 (chapter 1) things become more abstract / nicer to do math in. So I will work all odds in section 2, 3, ...
chapter 1 I mean
Manifolds are definitely an interesting topic. They're used in physics as you know
Configuration spaces for bots and spacetime fabric
I gave up on BananaCats for now. I'm leaning toward writing a Lean IDE if anything, but that math is sooo hard (type theory stuff)
22:03
0
Q: Duality of fields and duality classes

geocalc33Prescribe a space $Q,$with algebro-geometric objects (polynomials embedded in $\Bbb C^2$) $$ p_0,p_1,...$$ which form a ring $R[x]$ over the complexes. Actually, an algebraicly closed ring. Assume there exists a duality between $Q$ and $F,$ another space. Denote this as, $Q=\mathscr {dual}(F).$...

@EnjoysMath
It's at an a rudimentary stage in research. I like topos theory but I'm not sure that it would lead to a new / better implementation of "things".
Manifolds are kind of rad
in a geometrical way
@AlessandroCodenotti In undergrad I took a topics course on combinatorics of open covers which involved topological set theory (I think it was basically a topological set theory course with a specific focus)
Your notation is looking nicer / easier on the brain / more standard
THANKS
lol.
You helped with that a lot
so cheers
22:05
What's a field dual?
I gotta leave soon
I'm just asking about duality classes
of a field
@PaulPlummer sounds cool, are there notes by any chance?
What are those? Maybe give a link to your ideal reference wikipedia etc
@geocalc33 whoosh it's way over my head. But I might be able to start at a reference
22:07
okay
@AlessandroCodenotti Let me see if I can find some, I know someone who was typing notes during class and I think I got a copy at some point.
@EnjoysMath okay here's the thing. You know about initial objects in a category?
let's say you compose a space with initial objects
There exists a unique morphism from I to every other object and I is unique up to unique isomorphism
22:09
what?
THat's true what I said
what do you mean
about initial objects
what's I
ok
The initial object I
an initial object
22:10
anitial ya.
initial
okay so back to what i was saying
Prescribe a space of initial objects, (algebro-geometric objects)
I'm not sure how to do that
algebro-geomtric objects=polynomials embedded in $\Bbb C^2$
they form a ring
then take the ring $R[x]$
How does $\Bbb{C}^2$ fit into the picture?
22:12
hold on
Polynomials map $\Bbb{C}^n \to \Bbb{C}$
So $\Bbb{C}[x, y, z]$ for example
oh darn I meant $\Bbb C$
darn
Yes, change that lol
and thne...
$\Bbb{C}$ is like $\Bbb{R}^2$ in several ways
22:14
yeah facts
They're probably diffeomorphic but I'm speaking past myself there
okay and then take the ring of polynomials in $\Bbb C$ and assume there's a a transform, $\Psi$
which maps $R[x]$
What is $R$?
the ring of polynomials in complexes
follow?
In particular, maps from CC to itself should be isomorphic to maps from RR^2 to itself. It’s just that the kinds of maps one studies the former case may not look terribly nice when expressed in the latter case
22:17
@Semiclassical okay nice yeah
@AlessandroCodenotti Found them, not sure how understandable they will be as they where typed up during class (they look pretty nice though).
You would do this. Let $p_1, p_2, \dots$ be an at most countable collection of complex polynomials in the variable $x$. Define the ideal $I = (p_0, p_1, \dots)$ . Then in addition to being an ideal of $\Bbb{C}[x]$ ring, $I$ is itself a ring (all ideals are also subrings by def.)
Have fun writing out the RR^2 version of $z\mapsto z^3$, for instance
But since $\Bbb{C}[x]$ is a pid I think $I = (q)$, right?
(aka express that map in terms of real and imaginary parts. It can be done but I don’t want to.)
22:18
@EnjoysMath yeah I don't know if it's a pid
maybe
@PaulPlummer Nice, are they only somewhere or should I give you an email address?
(This is not to say you can’t work in terms of RR^2, just that it’s not as neat usually)
@AlessandroCodenotti You can email me at [email protected]
@geocalc33
In mathematics, especially in the field of algebra, a polynomial ring or polynomial algebra is a ring (which is also a commutative algebra) formed from the set of polynomials in one or more indeterminates (traditionally also called variables) with coefficients in another ring, often a field. Polynomial rings occur in many parts of mathematics, and the study of their properties was among the main motivations for the development of commutative algebra and ring theory. Polynomial rings and their ideals are fundamental in algebraic geometry. Many classes of rings, such as unique factorization domains...
I think that says somewhere that $K[X]$ if it's a Euclidean domain then it's a PID, and complex polynomials definitely have division with remainder algorithm, so form a PID. Not 100% sure though
@PaulPlummer Isn't it faster if I give you my address :P? [email protected]
22:24
@geocalc33 sent you an email with Dummit & Foote book attached. Check 8.2 it talks about PID's etc
and fields
I'm no expert, but that book will have the answer to that question mark
@AlessandroCodenotti Guess so, they should be sent
No, please don't send the missles :X
Did you know the US government is going into another arms race with Russia
Oh nice, there seems to be a lot of stuff on topological games, I've always wanted to learn more about those
22:26
"super fast missles" they kill us 7x faster
We didn't do any forcing in the class, as a chunk of us never had that level of set theory
It's about time to go to sleep here so I'm only skimming them quickly now, but I'll take a look more carefully tomorrow
Have a good one
Buono notte, demonic @Alessandro.
I've been doing nothing but forcing in the last two set theory classes I took so I guess that's a nice change haha
Hi @Ted
22:27
Bon apetit alessandro
Close enough @geocalc33
ABC
ABC
I have a question. What is the result of that limit in complex variable?

$\lim_{z->e^{i5pi/4} } z^{2/3} $
Use \lim
\lim\limits_{z \to ...} (...)
You need to talk about what branch of the 2/3 power you're using, @ABC.
ABC
ABC
This is my problem. I dont understand this. When I'm doing a limit in complex variable how can I understand the result? I know that a radix has more than one result.
in general
22:30
I don't understand why there's a limit in the problem.
Thanks for the notes @Paul looks like that was a really nice course! Where did you study for your undergrad if you don't mind me asking?
The limit from all directions must be equal by def of how they defined it
ABC
ABC
I want calculate limit
@EnjoysMath you changed your name again
I like it :)
If you have a well-defined branch of the function that doesn't have a branch cut along the ray in question, then the function is continuous, so who cares about limit?
22:31
THis is my original name, I also have enjoysmath @ google mail
That bananacats crap was sucking out my soul so I had to stop coding on that
Math is best done with pen / paper / chalkboard for now anyway
ABC
ABC
I'm doing a residue calculation and I'm using a "formula with limit". I have that complex function $\frac{z^{2/3}}{z^4+1}$
I keep asking you what branch of $z^{2/3}$ you're using.
You have simple poles, so you don't need this stupid formula for residues, anyhow.
The original problem has to specify a contour and a branch of the function.
@geocalc33 thx ;)
ABC
ABC
Wait I write contour
@ABC what complex analysis book do you use?
ABC
ABC
22:37
$A: se^{-it}, t\in[\theta,2\pi-\theta] \\
B: te^{i\theta}, t\in[s,R] \\
C: se^{it}, t\in [\theta, 2\pi-\theta] \\
D: te^{i(2\pi-\theta)}, t\in[s,R] $
@geocalc33 I would take out algebro-geometric and just write "polynomials"
So it looks like you're making a branch cut along the positive real axis. I still ask what branch of the function.
@geocalc33 I think I see where you're getting it incorrect
Presumably it's the principal branch, so you say $(e^{i\theta})^{2/3} = e^{2i\theta/3}$ for $0<\theta<2\pi$.
ABC
ABC
Yes I'm calculating a real integral
22:40
You've noticed an unfortunate nomenclature in mathematics. field in abstract algebra does not equal field in vector analysis / manifolds.
ABC
ABC
yes $0 < \theta < 2\pi$
The paper is clearly about the latter, but you're talking about the former !
The latter is closer to how physicists use it, though most physicists are more informal
At any rate, the limit is pointless here: You have a continuous function. You should understand where that stupid limit formula comes from (I never taught it or used it).
ABC
ABC
@EnjoysMath I'm using notes of my prof
22:41
(Typical physics usage is that vector/scalar fields are vector/scalar-valued functions on all of space)
A vector field is like the continuum of vectors pointing inward toward the Earth (an inverse square-law vector field). But field in abstract algebra are scalars (where as a vector field is a collection embedded in an ordered n-tuple of scalars). So you can't usually divide vectors but a field in AA by definition has inverse in both + and . ie. you can divide.
You CAN divide in $\Bbb{R}^2 \approx \Bbb{C}$ which is what you're noticing
now google quaternions, octonions etc. Turns out there's only a finite number of such ordered n-tuple structures that have divisonal propertiesi
I know the formula, @ABC. I just do NOT like it.
it's pretty helpful
22:43
You should understand residues from Laurent series. And you should know what the residue of $f(z)/g(z)$ is at $z=a$ when $z=a$ is a simple zero of $g$.
@Thorgott: It leads to students' memorizing and not knowing what the hell they're doing. So my experience of 40 years taught me.
ABC
ABC
I know definition of residue so that it is "linked" with Laurent series, but I don't understand how use the formula, i don't understand what is the result of the limit, because a radix has more than one "result"
It does not have more than one result once you've chosen a well-defined branch. You have to understand that before you start computing contour integrals and residues.
ABC
ABC
I saw the prof of that formula and I understand what she is doing
The fact that you keep insisting on the LIMIT shows me that you do NOT understand.
But I have experience with your stubbornness for several years now.
@TedShifrin Hello
ABC
ABC
22:48
I know that I don't understand something
(another qualifier: you do have physicists who use “fields” in the proper diff geo sense, but those tend to be either GR people or mathematical physicists)
ABC
ABC
I'm writing here because i need help
to understand my problem
I don't want to prove I know things
how large primes can we find if we let the computer work for 1 second? and how do we calculate such thing ?
of course that's true, but I feel like that goes for a lot of things
Which computer?
22:49
how do we generalize this to any time period
any average one
@JackOhara this post might interest you: math.stackexchange.com/questions/2962971/…
The algorithm is just adding two sets of integers elementwise
@EnjoysMath thanks ill check it now
ABC
ABC
So how can I solve?
ABC, they say if you ran into a hard problem there are a few easier problems (lower hanging fruit) that you can solve. I would look at the problems that build up to the result you're struggling with
I already told you ten minutes ago, @ABC. I wrote down the definition of the root in terms of the angle.
Hi @Jack
22:53
@JackOhara why generate primes?
@Semiclassic: There is an answer posted to that question. The answerer had a mistake the first time, but now I think it's right. But we end up with an ODE we cannot solve other than numerically.
There is a primality testing algorithm that is revolutionary solving a 100+ year old problem called AKS primality test. Uses 15 pages of beautiful abstract algebra & group theory to solve.
Written by 3 Indian students
@EnjoysMath i want to see the complexity theory asspect of it , also useful for RSA and crypto
what is the name of the alg?
oh sorry you wrote that
The algo paper is "PRIMES is in P"
it's a CS revolution paper. It's up there with FLT
22:55
thanks !
is this the fastest known method?
It will probably be a proof from the book one day
Well, it's not for generating
ABC
ABC
@TedShifrin Ok I understand now ... at least I think so. But I don't understand now, why if I do the sum of residue I don't get a real number but a complex number
it's for testing whether input $n$ is prime or not.
but isint Miller Rabin better that way?
It is not the fastest in practice, but it establishes an upper bound on deterministic polynomial time algorithms
@JackOhara probably
ABC
ABC
22:56
I think that my calculations are correct
There's no randomness in the AKS algo, if you run it again on exact same input, the path in code is exactly the same
But if AKS is deterministic that might be the best one
probalistic algorihms are scary haha
Well, it has notes at the end on how to speed it up as well as open problems that you might want to work on
You added up all four residues, @ABC?
And multiplied by $2\pi i$?
ABC
ABC
Yes
Residues are all in the curve
22:58
@JackOhara an open problem is what is the optimal complexity over all integer factorization algorithms
ABC
ABC
so I added up all 4
@AlessandroCodenotti It was at Boise State University. It was taught by Marion Scheepers who does research in these sorts of things
I personally would have tried a contour going around just one of the poles. Too much work your way.
optimal = minimal

« first day (3498 days earlier)      last day (1818 days later) »