The h Bar - 2017-12-12 (page 1 of 5)

« first day (2595 days earlier) ← previous day next day → last day (2333 days later) »

12:03 AM

LESS THAN A DAY TIL HATS!!!

it's the hatting time of season?

fabulous.

@Mithrandir24601 I would say that about The Raging Bull.

3 hours later…

3:45 AM

@skullpatrol Huh, already?

Time flies

Hi, everybody.

I solving all the problems in Lemons's random processes book

github.com/DanielSank/theory/tree/master/RandomProcesses/Lemons

4:35 AM

@DanielSank When life gives you Lemons, you solve all the problems in his random processes book?

4

4:53 AM

@BalarkaSen Correct

Are there any good books on quantum measurement theory?

5:19 AM

woah...

something in this code actually shuts off my computer...wtf...lol

unless my CPU is overheating or something...

funkyyy

@enumaris that'd be my guess

my CPU is liquid cooled though

=/

That doesn't necessarily stop it from overheating

I ran it quite heavily for 10 days straight doing data processing, I guess it still might be overheating heh

I'm pretty sure the issue is the loader is shutting down the que before the trainer has a chance to finish training

but I can't figure out why it's doing that

cus even getting rid of the que.close() command and the return statement completely didn't fix it lol

That I'll still have to get back to you on

5:35 AM

The trainer should take 731 steps per epoch, but when it errors, it has taken between 720 and 730 steps. I have diagnostic messages printed at every 10 steps and it errors after step 720. So essentially, when it errors it's about 10 steps from completion...which pretty much corresponds to the que size...

feels like it's gotta be the loader shutting down the que after it finishes queueing everything

You can check that by changing the queue size and seeing if the timing of the error changes in a corresponding way. Also you can print diagnostic messages at every step from 720 onward to make sure you know exactly where it's failing.

I haven't been able to find a pattern in when it throws the connection reset by peer error and when it throws the file not found error

oh, good idea

yep

after changing que size to 30

it errors after step 700

it was quite consistently throwing the error at step 720 earlier

error at 711 when que size is 23, so I guess it's not exact, but it's throwing the error at roughly when steps+que_size=total steps

smaller the que size, the later it throws the error

at que size of 10, error at step 724

so...yep...linearly proportional to que size

6:08 AM

welp...looks like the only hack work around I can get going at the moment is to manually break out of the loop at the last epoch after step 700...the program hangs (I guess the loader hangs)...but it will save the output and I can technically use the output later to continue training...seems like...a horrible work around though...lol

also now I have 1 print statement that is being printed 3 times to console...wtf...this is baffling...

6:25 AM

T-T

Cripes! -5°C in Chester.

user image

7:15 AM

We've finally got snow in London ;).

Though it's mostly melted now.

Captain Bohemian

do you take an umbrella to go out when snowing?

7:41 AM

Hello John, how are you?

@coderhk morning :-)

That question of yours:

You just need to calculate the potential energy

When the proton is at a position $x$ on the $x$ axis the PE is: $$-\frac{kQ_1e}{x} - \frac{kQ_2e}{x-3} $$

where $Q_1$ is the charge at $x=0$ and $Q_2$ is the charge at $x=3$

Calculate the PE at $x=5$ and $x=6$ and take the difference, and that's equal to the KE of the proton.

@coderhk I did the calculation to check it, and it does give the correct answer

@JohnRennie I see. Thank you very much. However, would the integration method work as well.

*?

Integration of the force just gives you the equation I quoted above for the PE, so yes it would work, if you did it correctly :-)

Ok. Thanks. :D

8:13 AM

mornin

8:55 AM

Hall's a nice book but Jesus so much introduction

I'm going through classical mechanics, which is a nice reminder but a bit long!

Starting from fields in QFT is literally sweeping so much stuff under the rug

what do you mean

QFT is from fields

that's what the F stands for

2

A: Quantum Field Theory is equivalent to QM of identical particles for free fields?

What we can say is that Quantum Field theory for free fields includes the relativistic formulation of free Quantum Mechanics of systems of arbitrarily many identical particles. Here, several well-known theoretical problems appear regarding relativistic localization and position operators. For l...

Yes you can do particles for free fields

But you can't for interacting fields

that's why you use fields

instead of particles

you can use particles for interacting fields but then you get something very very awful

You have to sum over graphs

9:11 AM

So apparently, and this may be a defect of the particle perspective, you can only look at interactions with no time dependence that pass systems from free initial particle systems into free final particle systems, not as a flaw but as a feature of the theory

"The momentum can figure in a consistent theory only for free particles; for these it is conserved, and can therefore be measured with any desired accuracy. This indicates that the theory will not consider the time dependence of particle interaction processes. It will show that in these processes there are no characteristics precisely definable (even within the usual limitations of quantum mechanics)...

... the description of such a process as occurring in the course of time is therefore just as unreal as the classical paths are in non-relativistic quantum mechanics. The only observable quantities are the properties (momenta, polarizations) of free particles. The only observable quantities are the properties (momenta, polarisations) of free particles: the initial particles which come into interaction, and the final particles which result from the process."

Is this all old stuff that fields overcame?

Pretty much

That's relativistic quantum mechanics

Which was what became QFT in the 30's or so

RQM is still used in some domains

Especially for very low energy processes

From this perspective you can derive quantum fields like immediately, but your starting point is just plain old QM and Relativity

where interaction doesn't matter too much

But right now I am denying the concept of interactions as anything but small perturbations, which is probably wrong

haha

Like in relativistic chemistry

Where you don't have to worry about particle production

9:17 AM

Yeah but this is also qft or at least qft as you'd find in Peskin, but with a RQM starting point but no gap getting to qft

My vague sense is RQM is basically postulating equations out of nowhere, like the Dirac equation as in Bjorken-Drell and waving your hands

RQM is fairly simple

It's the quantization process applied to relativistic equations

ie you quantize $$E^2 = p^2 + m^2$$

Which gives you the Klein-Gordon equation

which has some problems

Since $\pm \sqrt{p^2 + m^2}$ are solutions

and this translates to the quantum realm

You mean RQM is taking $E^2 = p^2 + m^2$ and setting $E = i \partial_t$ etc... as your quantization step?

pretty much yes

I mean you can do it more properly

Take the relativistic action $S = \int -m ds$

Apply quantization to the Poisson bracket

etc

But that's pretty much how it is

Basically the problem is antiparticles

Which doesn't matter if there's no interactions since you can't have negative energy states popping up without interactions

But otherwise things get weird

Alright, cool, well curveball - that is pretty distinct from what I am thinking of, but you might still call it RQM, where you literally start from the Schrodinger equation, but with an abstract Hamiltonian, and then derive the method of second quantization from it (e.g. the Walecka book), and the end result of this is getting quantum fields, the conjugate momentum, and even the quantum canonical commutation relations for fields, all starting from Schrodinger

then you need to perform some Things to get rid of problems

Well I do not know this method

9:27 AM

the only issue is the symmetry group you use to construct the explicit Hamiltonian - Galilean symmetry gives you the standard non-rel Hamiltonian, but due to relativity you need to go beyond Galilean symmetry, leading to reps of Poincare give you the qft cases as usual

The Dirac sea is a theoretical model of the vacuum as an infinite sea of particles with negative energy. It was first postulated by the British physicist Paul Dirac in 1930 to explain the anomalous negative-energy quantum states predicted by the Dirac equation for relativistic electrons. The positron, the antimatter counterpart of the electron, was originally conceived of as a hole in the Dirac sea, well before its experimental discovery in 1932. Upon solving the free Dirac equation, i ℏ ∂ Ψ ...

neat

@bolbteppa That wouldn't give you QFT

Dirac sea stuff is in Bjorken Drell as RQM stuff I never read because it's supposed to be wrong or something

Well I mean

What do you mean by "The Schrodinger equation"

If you mean $H \Psi = -i\partial_t \Psi$

Then yes, that is true

But that's because this is true for all QM

Yeah that

that's pretty much the definition of the Hamiltonian

but that doesn't really give you the dynamic of QFT

You don't get field operators from this

10:04 AM

So there is a way to start from Schrodinger for wave functions in the coordinate representation and end up with creation and annihilation operators, their commutation relations, quantum fields and the canonical commutation relations by transitioning to the occupation number formalism and a bit of combinatorics

mathb.in/20953

That's a quick write up of the transition in the bose case

Waleka book has the fermi case

yeah you can write free QFT as an infinite number of harmonic oscillators

So interacting qft, in the qft books I've seen, only does it perturbatively, which jibes with the whole explanation above that you can only measure free particle momenta and must ignore time dependence in interaction processes, only looking at the beginning and end free particles

Does it even make sense to talk about non-perturbative qft ("The momentum can figure in a consistent theory only for free particles; ... This indicates that the theory will not consider the time dependence of particle interaction processes. in these processes there are no characteristics precisely definable (even within the usual limitations of quantum mechanics); the description of such a...

... process as occurring in the course of time is therefore just as unreal as the classical paths are in non-relativistic quantum mechanics" seems to say it does not make any sense)

I've been living under the delusion that qft being perturbative is some gigantic massive flaw, when apparently it's at least partially characteristic, which is crazy

You can do non-perturbative QFT

It's just not very pleasant

or possible, in most cases

You can do it okay in one dimension

I think it's also mildly possible in 2 dimensions

and for some specific theories in 4 dimensions

Like very nice supersymmetric systems

I guess I'm just thinking out loud as to whether that whole subject is just a big error, it's all math with no physical content, or whether these passages are old and flawed, because those passages seem to say you can't do it

It's probably written with a very specific model in mind

10:16 AM

I keep reading the claim all of qft is axiomatically non-perturbative 'by definition/axiom', physics.stackexchange.com/a/33207/25851 which if you take those passages above literally, is a gigantic error haha

There's like a whole bunch of QFT axiomatizations

And none really work that well for the standard model

"standard" QFT isn't really an axiomatized theory

It's just kind of a loose bunch of ideas that work

it's a sausage factory

Just don't look too close how it's made

and keep eating

That's part of the charm

QFT is basically a strings and tincan theory

I mean there's some rigorous QFT, but for the most part it's not really used

user228700

Hi, everyone :-)

Hello

10:26 AM

28

A: What is the issue with interactions in QFT?

There are many different problems with interactions; or, rather, many manifestations of the same problem. For example, interactions are always non-linear in the equations of motion, e.g., $$ (\partial^2+m^2)\phi=\lambda\phi^3 $$ or similar equations for QED. As the operators are distributions, ...

Finally, a vague appreciation of Haag now that I vaguely know what it says, sync's well with what's written above!

@Kaumudi.H morning. Should I ask how the exam went?

user228700

Yes, it was alright, thanks :-)

user228700

How has your day been?

@Kaumudi.H cold!

user228700

Ah :-/

10:28 AM

4 hours ago, by John Rennie

user image

But that's OK. I'm nice and cozy indoors staring out at the snow and ice :-)

I'm at work in my parka rn

They're not very good at heating the office

user228700

@JohnRennie :-)

user228700

@Slereah Oh, wow.

user228700

@JohnR: I am hungry and bored :'-( Quite tired, too.

@Kaumudi.H it must be lunch time in Kochi. Grab some lunch?

user228700

10:32 AM

I will go out and have a grilled chocolate sandwich in just a bit but there is nobody to go with and I am starting to get tired of my own company, haha.

"Annual temperatures range between 23 and 31 °C"

I'm guessing @Kaumudi.H has less troubles with the cold!

@Kaumudi.H If I'm bored and a bit fed up I quite like to go for a walk. I find the (gentle) exercise cheers me up, and I can relax into a daydream as I'm walking.

user228700

@JohnRennie No, I have tried that before and it doesn't quite work on me.

user228700

@Slereah Yep, I'm quite comfortable living near the Equator :-)

Go to the library? Spend some time browsing?

user228700

10:34 AM

What library? :-P

There must be a library in Kochi ...

I saw that typo :p

Don't get too cocky @JohnRennie

@Slereah Bit of a kock up :-)

user228700

@JohnRennie Hmm, there is, but it's quite a way away and my cycle is not very good to ride in hilly areas like these.

Mithrandir24601

@JohnRennie ::groans:: I'm just leaving the ~20 degree warmth to return to that... This is the downside of going to the tropics in midwinter

10:37 AM

@Mithrandir24601 That was a short stay. When do you leave?

Actually your timing is good as it's going to get significantly warmer over the next week.

@Kaumudi.H does the college provide any recreational facilities, e.g. a gym?

user228700

x'D Nope!

user228700

Man, I really do appreciate your efforts to suggest an activity for me!

There does seem to be a big difference compared to UK universities, where the social life is a large part of the experience.

Mithrandir24601

@JohnRennie in about 5 and a half hours - I'm actually already on the bus to the airport. Would have really liked to stay longer, but it was too last minute. Also, being in China is vastly easier if you know someone from the area :)

Is that generally the case in India or is MEC particularly bad?

user228700

10:41 AM

@JohnRennie I...don't know, actually.

I guess we need some of the other Indian students to comment, but no-one is around ATM.

user228700

MEC is particularly bad, absolutely, but I can't be sure about the social life in other, better colleges, such at the NITs and IITs.

user228700

OK, I will do what I have been doing for so many weeks now-listen to this week's episode of "Dear Hank & John" on my walk to the "cafe" and back.

@Kaumudi.H Enjoy the walk/podcast/sandwich in whatever order is appropriate

user228700

Thanks, John :-)

10:47 AM

You could install Python ...

Mithrandir24601

@JohnRennie to be fair, this varies a lot in the UK as well - I remember Nottingham as not really having any social life, outside of the people I knew from the course and from outside the uni

Or set up Google drive ...

@Mithrandir24601 really? Cambridge was crawling with student clubs and societies.

Mithrandir24601

@JohnRennie so is Bristol :) but Nottingham just felt a bit dead really - the campus is lovely and I got on well with everyone and everything, but the student life there just seemed non-existent. I'm not into clubbing or anything, so I don't know if that made a difference or not though

And I can confirm that the Chinese uni I was at these past few days has a gym, badminton, tennis etc. Etc.

11:11 AM

Ohai

11:37 AM

So, we're going to be checking to see if Oumuamua is an alien spaceship ...

Anonymous

@JohnRennie We have lot of sports clubs here. Also literary club, science club(which is a robotics club in disguise :P) and a coding club. There is a so-called gym which no one bothers to use :P. It might shut down in a few years. I don't participate in most of the events other than the friendly football matches. I hope I can convince people to make a "serious" science club here which does real science and not just discusses popular science.

.. because well you never know :-)

Anonymous

Hehe. BTW the number of clubs varies widely depending on the university. All the well established well funded old universities in India have a lot of clubs and social activities. The newer ones: not so much.

@Blue ah, OK, that sounds much more like the experience I had at university.

Anonymous

MEC is relatively a newer college. So what Kaumudi said is expected. I'm sure over time it's gonna develop. It's surely one of the best in Kerala.

11:40 AM

To be fair MEC isn't a campus university, so it will be limited in the facilities it can offer. For example I'd guess a gym would have to be away from the main buildings.

BTW how is the laptop doing now it's had a chance to settle down?

Anonymous

Ah, that is unfortunately true. They also seem to have very strict rules like no mobile phones in class and uniforms...aaaah! :p

Anonymous

@JohnRennie AWESOME

Anonymous

:D

Cool :-)

Anonymous

In waiting for 14th...I'll get back home and ping you for the dual booting :)

Anonymous

11:44 AM

Damn...last exam is humanities...XD....didn't study it the while year

Anonymous

Now I've to be in maximum overdrive and complete 6 months worth of syllabus in 1 1/2 days

@Blue at your convenience. You'll need the Ubuntu install image. You can either grab it from the Ubuntu web site or I have a copy I can put on my server. It's about half the size of the Windows install image :-)

Anonymous

@JohnRennie Sure, we'll see

11:56 AM

Ew that paper writes "coördinates"

How 19th century

Mithrandir24601

@CooperCape nihao!

12:55 PM

@EmilioPisanty for what is worth (and what you may care), upon further reflection about yesterday's discussion, I find myself mostly agreeing with you. There really isn't much of a point, at the present stage, in having (trying to have) a separate site like the one of that proposal.

> QFT evades Haag's theorem through renormalisation, and only because the latter is an intrinsically ill-defined operation.

This is soooooo (insert word)

1:17 PM

@glS I'm neither for it nor against it, to be honest. I am against it being done wrong, but that's a different matter.

anyways

each their own and all that

btw, this

1

A: How can we certify that the randomness in the measurement outcomes is not due to randomness in the state preparation?

How do we know that the randomness is not caused by the state preparation itself? It depends what you mean here. In some sense it is caused by the state preparation: the "state preparation" generates the quantum state, which is the cause of the randomness in the measurement outcomes. But ...

was one quality answer

I started writing an answer about three times and gave up thinking that it was just going to be too much trouble

but that answers says most of what I wanted to say

@Slereah the New Yorker consistently oös their oöable oos, if I remember correctly

It is good that acuriousmind and other quantum guys are currently on, cause later on I might need to consult you guys on a philosophy of quantum mechanics problem raised by a philosopher I chat to today in a conference

@EmilioPisanty yes that's what I've thought. At the present stage, also having a look at the people that seem to be there, it's just probably not going to go well. And there isn't much of a point in doing it if it doesn't go very well, because really physics.se shouldn't be underestimated.

@EmilioPisanty thanks, I appreciate it

1:44 PM

@Mithrandir24601 I don't speak these foreign terms.

I think it's some weeb term.

0

Q: Are the components of 4-vectors the physically measured quantities?

I am very confused with the difference between components of four-acceleration and coordinate acceleration. If I was in an inertial frame observing an accelerated object I would say its four-acceleration is $$ a^{\mu} = \Big( \gamma^4 \frac{\mathbf{v} \cdot \mathbf{a}}{c},\gamma^2\mathbf{a}+\gam...

special-relativity

measuring the time component will be hard

@BalarkaSen aret thuo arundeth?

"For any bounded operator A, there is a unique bounded operator A∗, called the adjoint of A"

Are adjoints defined for unbounded operators?

Yes, but the construction is nontrivial

1:48 PM

(I am reading Hall)

I assume he talks about it later

Yes, he defines it in chapter 8 or so

I'm chapter 2 so it might take a while

@ACuriousMind Dude

do you want to play BL2 today/tomorrow?

I'm done with school after today

@Slereah if I say "Merry Christmas" in an email am I Christian ISIS?

is that considered a bad thing now

Only in your crazy country

use the standard American way

Krusty the Clown - Merry Christmas

@0celo7 I'm busy both evenings my time, but could do tomorrow afternoon (no idea what time that's for you)

1:58 PM

@ACuriousMind works for me

Can you give me a UTC time?

I'm staying in town for a few days to wrap up some business before going home, so I'll be here tomorrow morning

Hmmm...sometime after 1200 UTC and before 1800 UTC, I think

Time to stop chapter 2 and investigate appendix A

« first day (2595 days earlier) ← previous day next day → last day (2333 days later) »

Transcript for

Dec '1712

The h Bar

General chat for Physics SE (physics.stackexchange.com). For M...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile