The h Bar

12:13 AM

I want to sort of discuss how renormalization works in $\phi^4 $ theory

2:17 AM

Is there anyone here who is familiar with the linked cluster expansion method? I'm confused how to calculate the lattice constants - which is defined to be the number of ways a cluster can be embedded in the lattice divided by the total number of sites.

For example, for corner-sharing squares, (cf researchgate.net/figure/…), I don't get why for c=2, the lattice constant is 1/2.

Cows

4:46 AM

So I currently online now

I think tonight, I will understand this re-normalization thing once and for all.

I have explored this idea from a few angles recently, but tonight I think I want to attack it as it is used in $\phi^4 $ theory

rob

@Nat Whenever I have a student who makes an order-of-magnitude mistake like that, I offer to loan them a dollar.

Cows

So one starts with some lagrangian. . . and I am shooting from the hip here and so

and ends up with the well known eom with $\lambda \phi^3$ and $\frac{1}{3!}$ terms and so forth

some differential equation

One then applies Greens technique and so forth

Let us skip the fourier decomp business for a bit

now

there is something that has that $\frac{1}{Q}$ form which gets some $i \epsilon$ thrown that then gets pathological, and needs regulating and so forth

let me look up the form on google. Give me a sec

right so the $\frac{1}{p^2 + m^2}$ thing integrated over $d^4p$ you know the deal . . .

so far nothing exotic going on. . I mean some regulator can be thrown in or something

then a taylor expansion and redefinition is made

the logic gets hazy in some accounts of the next things that go on

I have not decided to attempt this stuff in a while,

I do believe that nothing exciting is happening here, one just taylor expands things of interest and then uses fancy words and new terminology to redefine the couplings et al

The formal details as presented in qft txts are redic to be honest

I mean , . . .

enumaris

no idea what u talkin bout bro

Cows

I am trying to make sense of renormalization

in the simplest possible case but not ising model

enumaris

5:01 AM

o.O

what's the ising model have to do with renormalization...

Cows

as for ising model one can just play games on triangular latices and redefine spins

by some clever averages

I've simulated things like this with ease

ok . . . . well I struggled, but i have simulated this stuff before

well , let us discuss ising model a bit

It ,. . . as you know is just a model of spins on a lattice

neighbors interact, and so forth given by a formula

one can define a new spin as average of some spins, in a certain way depending on your lattice, and so forth a

and put it on a computer to simulate . . . you've done this before right?

They use similar language and so forth

enumaris

o.o

Cows

@enumaris am I making sense ? You've done this before no?

enumaris

I haven't worked with the ising model

Cows

oh, well I've written some simulations in the past. Don't worry anyone can do it . It is not impressive lol

enumaris

5:08 AM

I'm having trouble figuring out how to initialize my neural net in pytorch...-.-

Cows

What is hardcore is not just running the renormalization algorithm for a (very simple) qft but knowing what it means. This has evaded me for ~6 years lol .

I mean I can't even understand $\phi^4 $ theory to a satisfactory level

enumaris

I dunno much about renormalization

Cows

I finally only figured out some basic things

well , I am thinking I actually know what it means in some simple cases, but I want to talk to a few people, just so I am confident I am not kidding myself

I mean, there is a whole math language that describes this stuff that I first explored some time ago, without success, but I think it makes sense now.

At least for a very simple theory

Before I explore it in that context, my goal is to just try to convince myself that I understand the gymnastics in physics rather well

It seems to me just a cute game of talor expansion and redefining things in terms of other things

This is a load to say but I have a cocal project where I put my money where my mouth is and actually do this ,

but again it is good to have people check this stuff out

@enumaris you've taken qft no?

enumaris

yeah, only the first quarter tho lol

Cows

That is great ! I have not officially taken qft. My first in class setting was sitting in a second semester qft class at a university which shall remain unnamed :P

but you are experienced with scalar field theory no? I guess these are the simplest possible theories or something

This is about all I am confident I will ever really understand :P

well I mean at least in the short run

Everything else is outrageously hard to understand for me

I mean I can run through the mechanics if sufficiently motivated, but understanding hehe no shot yet

oh well

I will try to work on cocalc a bit.

It is difficult to formulate thought and latex here . hehehehe I probably sound incoherent heheh :P

I will share a doc with you when I formulate some trivial things that are well know, so you can may be check me on it

I will be highly trivial type first day, of first semester qft stuff mostly.

@enumaris are you coding right now?

enumaris

5:27 AM

kinda

but it's not really working lol

Cows

hehehe

enumaris

working on a library for reinforcement learning

but the neural net is not working as expected ._.

eventually this library will create AGI...muahahahahaha (not really)

Cows

6:19 AM

Hehehe AGI is the big goal hehehe

Anonymous

7:37 AM

mornin

Anonymous

seems quiet around here

Sid

Afternoon.

Anonymous

sup?

Sid

About to eat chicken. :P

Anonymous

Lol. Where are you now?

Sid

7:44 AM

@blue what about you?

Anonymous

@Sid Listening to random songs and reading a few things

Sid

Reading what?

Anonymous

Seth Lloyd's ML papers

Anonymous

You were in a summer training program (on programming), no?

Sid

Yeah. Still in. Today's a holiday because Sunday. Sort of programmung, yeah.

Also 3D Graphs are insanely cool. XD

Anonymous

7:47 AM

3D graphs?

Anonymous

What sort of?

Sid

Nothing much really. Just looking at them. We will probably use those all enough.

Anonymous

GeoGebra has a crappy 3D grapher. Maybe you could come up with something better: like Desmos 3D ;)

Anonymous

Actually that would be a brilliant summer project

Sid

Hmm. That sounds like something to try. But, I am having stuff to do. >.<

Anonymous

7:51 AM

Hehe...I can understand :P

Anonymous

@Sid What are you having to do though?

Anonymous

Python coding stuff?

Sid

8:14 AM

Learning MATLAB. And trying to start learning R on my own

Kenshin

Hello there

Anonymous

@Sid You might enjoy this then. It's in R. The textbook is this: web.stanford.edu/~hastie/Papers/ESLII.pdf

Kenshin

Can someone please help me with my question on knowledge

Sid

@Blue I will see that when I get on a laptop. Thanks.

Slereah

9:29 AM

Hey chaps

Anonymous

Holla

ACuriousMind

ahoi

Balarka Sen

this chat ded

Yuri Gagarin - At The Center Of All Infinity (2015, Full album)

►

John Rennie

9:47 AM

Not dead, just sleeping :-)

It is, after all, a Sunday morning.

Balarka Sen

That does seem to explain the anomaly

@Slereah Jets are chasing me

Slereah

Fighter or bundles?

Balarka Sen

Both

These are not ordinary jets. There are k-jets

Slereah

for what $k$

Balarka Sen

Arbitrarily large

Gromov's driving

2

Anonymous

10:09 AM

@BalarkaSen sounds legit

Balarka Sen

straight from the depths of /sci/

Novalium Company

Guys, a second integral is just taking the integral of an integral, right? Just like a second derivative is taking the derivative of the derivative?

ACuriousMind

@NovaliumCompany "second integral" is not a common name for that

You might find "second anti-derivative", but I haven't heard that all too often, either

Novalium Company

Ok, then what is this $$\iint$$?

Anonymous

Multiple integral

The multiple integral is a definite integral of a function of more than one real variable, for example, f(x, y) or f(x, y, z). Integrals of a function of two variables over a region in R2 are called double integrals, and integrals of a function of three variables over a region of R3 are called triple integrals. == Introduction == Just as the definite integral of a positive function of one variable represents the area of the region between the graph of the function and the x-axis, the double integral of a positive function of two variables represents the volume of the region between the surface...

Novalium Company

10:21 AM

So the multiple integral can help you find volume?

Anonymous

Sure

Novalium Company

Hmm, interesting.

John Rennie

@NovaliumCompany it's common to integrate things twice or even three or four times. Typically we do this when a function is a function of several variables.

ACuriousMind's point is that we wouldn't use the term second integral to describe this.

Novalium Company

Double integral?

Multiple integral?

John Rennie

Possibly we'd call it a double integral.

Novalium Company

10:23 AM

So using double integrals I can find volume, that's nice. What about triple integrals :D? Find 4D volumes xD

John Rennie

A double integral would give you area not volume.

A triple integral would give you volume.

Novalium Company

I thought $\int$ would give you area?

John Rennie

The infinitesimal area $dA$ can be expressed as the product $dx dy$. This is just area = length times width in infinitesimal form.

Then the total area would be given by $\iint dx dy$

Novalium Company

Area can be found using a single integral, no?

John Rennie

Likewise the infinitesimal volume is $dV = dx dy dz$

So $V = \int\int\int dx dy dz$

Novalium Company

10:27 AM

Wiat wait. $\int$ gives you the area, doesn't it?

Anonymous

@NovaliumCompany Ah, that's the area under a curve. We're talking about finding the area of (say) your computer screen.

Novalium Company

Ohh, got it.

But it's a rectangle?

B*H

Why use integrals?

Anonymous

Because if it gets dented it will no longer remain a perfect rectangle ;)

Novalium Company

Ok, let's say I want to find the area surface of a lake.

Anonymous

That's better

Novalium Company

10:30 AM

How would I do that using a double integral?

Anonymous

youtube.com/playlist?list=PLwumBuNUWTLLCI9CDXeypUm6u1fgjnSK0

Novalium Company

Ha, thanks.

Question: $\frac{dv}{dt}$ of a velocity-time graph, will give us the acceleration and $\int v(t) dt$ will give us the distance travelled right? So can I view derivatives as a form of a weird division and integrals as a weird multiplication?

ACuriousMind

@NovaliumCompany Well, both are limits of "infinitesimal" divisions/multiplications, but I don't find it particularly helpful to think about them that way.

The Leibniz notation $\mathrm{d}v/\mathrm{d}t$ is suggestive, there is no actual division of some objects $\mathrm{d}v$ and $\mathrm{d}t$ taking place

Novalium Company

I know I know.

But v/t = a?

John Rennie

@NovaliumCompany ACuriousMind is of course quite correct, but physicists treat $dy/dx$ as a fraction all the time and we (mostly) get away with it :-)

ACuriousMind

10:41 AM

Yes?

@JohnRennie I do allow physicists these shenanigans, but I think one needs to first understand why it's wrong before one does it anyway :P

Novalium Company

I know that $\frac{dv}{dt}$ dosn't represent division, is that what you are saying?

Anonymous

@NovaliumCompany What is v/t = a?

Novalium Company

$a = \frac{\Delta v}{\Delta t}$

Anonymous

Better

Novalium Company

:D

Anonymous

10:45 AM

@NovaliumCompany But that only works when the increase or decrease of velocity is constant per unit time

Novalium Company

I know, that's why we use derivatives?

Anonymous

What you wrote is actually the average acceleration

Novalium Company

Yep, and derivatives help us find the acceleration at the exact point in time, my mistake sorry.

Anonymous

@NovaliumCompany For finding instantaneous acceleration you have to use $dv/dt$. However, instantaneous acceleration = average acceleration in certain cases

Novalium Company

btw, I've seen integrals that don't have like from 2 to 5 for example, how are those called and why are they that way?

ACuriousMind

10:55 AM

@NovaliumCompany You mean $\int_a^b$ where the $a$ and $b$ is missing?

Novalium Company

nvm, i found it, please ignore this ^

Indefinite integrals is what I was looking for, thanks :D

Novalium Company

11:10 AM

Btw if I graph the base and the height of an object. For example the b(h) = ... , then the first integral is the area?

Novalium Company

11:21 AM

$\sqrt(x) = x$ on the 1/2 power? wut, I thought it's equal to $x^2$?

ah. nvm, ignore this please ^

Ooker

2:09 PM

Hi @knzhou. Following the last comment, what do you think about it? I ask you in chat first because I guess the answer ultimately boils down to "that's because how we measure it". I also think that it's because of the symmetry groups with the Perter-Weyl theorem.

I want to check whether the answers are more complicate than that or not before really asking the question. Also this question may be interesting to you: What is the significance of differential operators over other operators in group theory?

Balarka Sen

We see ODE's because we study the evolution of dynamics of a given physical or biological or chemical system over time.

If you model any given such dynamical system as a space $M$ with an evolution operator $f_t : M \to M$ that is time ($t$) dependent, the natural question is how the evolution changes with respect to time. The answer to that is an ODE $\partial_t f_t = X$

Sir Cumference

2:30 PM

@NovaliumCompany I'd really recommend not looking much into multiple integrals until you strongly understand single variable calculus

It's a completely different world and can end up confusing you in the basics

navnav

Hi, I've been reading about the interactions between light and matter. I've come across this video, and it's completely thrown me off. I'm not sure if the author actually has a point, or is just talking nonsense.

He claims that light is a "coaxial circuit", and that QED & wave particle duality are "Quantum Quackery".

Is there any sense to this video or should I just ignore it?

He refers to explanations in another video, which may be this: youtube.com/watch?v=_k1Vsqn2cuw. In this video, he claims that light doesn't travel at all and that the absorption and re-emission process is what 'crazy people' believe (amongst other things!)

This is the original video: youtube.com/watch?v=qpcQNYa3VZ8

ACuriousMind

@navnav Just from the description it sure sounds as if you should ignore it :P

navnav

what's really frustrating is that this video appeared on the first page on YouTube for "single slit diffraction experiment"

ACuriousMind

Well, YouTube is not exactly the best place to get your scientific information from...

Slereah

@ACuriousMind I get all my infos from Alex Jones

3

navnav

2:41 PM

I was just looking for a video on the single slit experiment.

ACuriousMind

Why a video? It really strikes me as the most inefficient way to convey technical information, but maybe that's because I read rather fast...

navnav

Well, I wanted to see it in action, preferably where they change the slit width to see how doing so will affect the interference pattern on the screen.

If you can suggest a better site to use, that would be great.

Balarka Sen

@Slereah So you believe in the interdimensional vampire child molesters huh

lılostafa

@navnav I suggest Wikipedia as the starting point. (Also, ACM is busy watching his team vs Mexico match right now)

Slereah

@BalarkaSen it just naturally follows from string theory

Secret

3:01 PM

I don't disagree that I agree

ACuriousMind

→ 2 messages moved to Trash

navnav

Thanks @lılostafa, I've already had a look at the Wikipedia pages on diffraction and the Huygens-Fresnel principle, which has enabled me to gain an understanding of the mechanics behind diffraction. But in this case, I just wanted to see it in action.

After reading http://www.alternativephysics.org/book/Diffraction.htm, which focuses on the CED understanding of diffraction and mentions some problems with the classical understanding/Huygens-Fresnel principle, I stumbled across the video and it just threw me off

navnav

3:30 PM

wait, is that site I linked to also a BS Physics/flat-earther site?

Gosh, it seems like I'm looking for them, but I'm genuinely not.

I ask because of this forum post: physicsforums.com/threads/is-this-gr-explanation-legit.513491

I know the name is 'Alternative physics' but I just assumed that meant Classical physics

bolbteppa

@navnav crazy website

navnav

So I guess what I read on that page, RE the mechanics behind light bending from diffraction (in 'The details' section) is incorrect? It states that the light bends because the energy of the light waves excite the electrons at the edges of the slit, without being absorbed, which cause them (the electrons) to re-emit waves at the edges, in all directions. Is that all nonsense?

ACuriousMind

3:50 PM

@navnav Yeah, that's nonsense. There doesn't need to be anything capable of emission in the material of the slit in order for diffraction to occur - note that water waves diffract on obstacles, too.

Your description sounds as if something took the mathematical content of Huygens' principle - that every point of a wavefront can be treated as the emitter of a spherical wave - too literally.

navnav

OK, thanks, but then why are there interference patterns on the screen for a single slit experiment?

On that site, they claim the interference is caused by the waves emitted at the edge of the slits

There are fringes on the screen, which suggest multiple sources. What causes them? Feel free to just send over a link that explains this (don;t want to take too much of your time!)

ACuriousMind

@navnav By Huygens' principle, you can treat every point of the slit as the source of a spherical wave. The Wikpedia articles are actually pretty decent: en.wikipedia.org/wiki/Diffraction, en.wikipedia.org/wiki/Huygens%E2%80%93Fresnel_principle

navnav

OK, but why only at the slit. Why don't light waves bend when travelling through empty space? What makes every point of a wavefront a new emitter when it travels through a slit? The wave is not interactinbg with any particles, so what is emitting the new waves?

ACuriousMind

@navnav The same is true for a wave in free space.

If you actually think about it/do the math, you find that Huygens' priniciple for a spherical or planar wave just leads to that wave spreading further.

The slit now "cuts away" part of such a spherical wave - and that's why you see a different behaviour after it, since your starting condition is no longer a full spherical or planar wave.

navnav

4:05 PM

If every point of a wavefront in free space is a new emitter, how is a beam/laser possible? How can coherent light work?

Wouldn't each point emit in all directions?

ACuriousMind

10

Q: Huygens wave theory not applicable to lasers or parallel beams of light?

According to Huygens wave theory, every point on a wavefront acts as a secondary source of waves. Using this principle we can never have pretty narrow parallel beams of light right? Like lasers? There will always be "leakage"(wrong terminology?) of waves from the edge of the beam right?

navnav

Looks like I've got some reading to do. Thanks @ACuriousMind

navnav

4:27 PM

Just one last question: With refraction in the manner of Huygens, the points that are emitting the refracted light are the particles that make up the material, right?

My understanding is that the light wave excites the particles that make up the material, which results in re-emission. So when a light wave is incident on a transparent material, and the indecent angle is not 0, the light wave will interact with the particles in a sequential manner, which results in a phased-array like effect, which is why the refracted light bends at an angle from the normal.

So in other words, this image taken from the Wikipedia page on Huygens–Fresnel principle: en.wikipedia.org/wiki/Huygens–Fresnel_principle#/media/File:Refraction_-‌_Huygens-Fresnel_principle.svg

Do the points in that image represent particles of the material?

Link isn't working. It's the first image on https://en.wikipedia.org/wiki/Huygens–Fresnel_principle

Kelthar

guys, do you know how to derivate this with respect to phi? Aren't we supposed to condier phi and phi dagger independent variables?

Sid

4:53 PM

@ACuriousMind You won't be hearing about today's game in elevator or office anymore. :P

lılostafa

damn
My prediction was Germany 3-1 Mexico

Balarka Sen

Nerds

Loong

German teams traditionally start slowly.

Sid

@Loong 3 out of the last 4 defending champions have been knocked out in the group stages. Signs don't bode well for the Germans

Avnish Kabaj

Germany will win

Oh noes

0-1

How

lılostafa

5:07 PM

@AvnishKabaj Mexico's formation was 5-4-1 in the 2nd half :)

Sid

@AvnishKabaj That's the World Cup for you

@lılostafa Mexico built the wall. :P

lılostafa

:)

Sid

Unfortunately, the Germans paid for it.

Avnish Kabaj

@Sid nice

I see what you did there

Cows

Can someone walk me through renormalization in $\phi^4$ after throwing in lambda and regulating?

ACuriousMind

5:12 PM

@Sid Yay, I guess?

Cows

I want to compare that explanation with what I think is happening

I think it can be sort of like 3 or 4 steps

ah life . . . . . may be I should just do it myself . . .

ok fine! I'm going to do the simple calculation and compare. The problem is that I do it slightly differently

Let me make some tea first

Today should be an interesting day of random qft computations all day. . . . . paper and pencil in my room . . . .

I know words like real scalar or complex scalar and quartic interaction etc are the types used. . . . I am deliberately just typing what comes to my . . . . at any rate. I am going to sit my laptop on my desk (still in bed . . . . just woke up); make some tea, and have it with renormalization for breakfast.

Making tea will take too much time . . . . Going to just bang out this stuff instead

let me first play some classical games

you know . . . . so I can actually write some rigorous things (hehe)

Eulb

5:41 PM

@ACuriousMind says the dude who says he never reads books

Mithrandir24601

6:02 PM

@Eulb The file just gets uploaded (in binary, converted from ASCII) directly into the 'brain', evidently

ACuriousMind

@Mithrandir24601 ASCII? I'm a modern AI and use Unicode!

lılostafa

NEW: Agency that monitors earthquakes in Mexico said an “artificial quake” was registered in Mexico City, possibly caused by “massive jumps during the goal from the Mexico national soccer team.” Mexico upset reigning World Cup champion Germany 1-0. https://abcn.ws/2t5wez8

Tweeted by ABC on June 17, 2018 at 6:01 PM

^Nonsense?

Mithrandir24601

@ACuriousMind beg pardon. I just figured that ASCII might take up less space, giving you room for twice as much stuff

ACuriousMind

@lılostafa Depends what strength they claim this "quake" is. Reminisicent of what-if.xkcd.com/8

Loong

@ACuriousMind wdr.de/tv/applications/fernsehen/wissen/quarks/pdf/…

ACuriousMind

6:11 PM

@Loong Hmmm...they got 0.2 Richter with ten thousand people being told to jump synchronously. Depends on how coordinated the Mexican jumping was then, I guess

Loong

Quarks&Co Rock am Ring Erdbeben

►

Emilio Pisanty

6:27 PM

@lılostafa @ACuriousMind It's true. Ditto for the surge in the tear meters all over Germany at the final whistle.

Bob

Hi

would you anybody care to look at my question at the following URL:

-1

Q: Two Bodies in Space

Below is a problem I am working on. I believe my solution, so far, is correct. I plan to complete the problem by using numerical techniques to produce a good estimate of the answer. Is my solution so far, right? Thanks, Bob Note: This is not a homework assignment for me. Problem: At time $t = 0$ ...

homework-and-exercises newtonian-mechanics gravity

ACuriousMind

@Bob I'm afraid we don't do "check-my-work" type questions here.

Bob

should I delete my post?

ACuriousMind

Not necessarily - you could edit it to ask about a specific concept you're unsure about instead of just asking about correctness, for instance

Bob

would it okay to ask how to proceed?

ACuriousMind

6:37 PM

Without any more specific question, not really

You could also have a look at physics.meta.stackexchange.com/q/391/50583 for places where such questions might be on-topic

Bob

I thank you

have a nice day

ACuriousMind

You too! Thanks for being understanding :)

Anonymous

That ToughStem site mentioned in that Physics SE answer is sort of creepy. I had asked around 2 questions there around 1 and half years ago, and till date, there's no comment or any other activity on it. I suspected that might be because there are no active users on the site. However, from the main page, it seems the Hot Questions get viewed almost every minute (which is weird considering I've never seen any other live activity on that site)

Anonymous

All the posts there seem to be a collection of Q&A's from 2015-2016

ACuriousMind

That's valuable information - maybe add it in a comment to the meta answer?

Anonymous

6:55 PM

@ACuriousMind Done

Anonymous

The ToughStem site doesn't seem to have active users anymore (I had posted 2 questions there around a year ago, but got absolutely no response). It seems more like a collection of Q&A's posted in 2015-2016 (probably by the same person who created the site). However, what I find sort of creepy, is that, in spite of no sign of live activity (like commenting/answering), the Hot Questions on the main page seem to be "viewed" almost every minute. Could some clarify what's going on with that site? Is it using some automated process to "fake views"? — Blue 2 mins ago

Cows

7:25 PM

Just got back from a walka. Bought some paper for all the power series expansions and plots I'd be making later on. Now going to cook and eat something and eat. Should be quite fun! Going to try to understand some typical physics algorithm

Quick question. If I write an epic real honest to God well put tutorial on renormalization would you guys appreciate it fiscally ;-)

Also, . . . . I will be sharing some important news in the next 2 weeks if all goes well . .

At any rate let me eat something and sit here and play some math games for a bit

do ping me in the process because I am bored . . . . super bored . . like . . . . . .

bolbteppa

7:42 PM