The h Bar

user280247

18:10

Is there any easy way to see that buoyant force is shape independent?

user280247

I mean, it is easy if we admit archimides principle, but how did he realize it?

enumaris

By shouting Eureka in his head over and over until he got there?

user280247

o_0

user280247

I think possibly he didn't say anything about buoyant force but only about equal volume displacement

Emilio Pisanty

18:31

@santimirandarp well, you can calculate the force explicitly and show it comes down to the volume

Semiclassical

the origin of various principles/concepts from antiquity is, alas, one of those things which lies beyond the veil of history

Lozansky

@Semiclassical Hi Semi :D

Semiclassical

though Archimedes' "On Floating Bodies" does survive: en.wikipedia.org/wiki/On_Floating_Bodies

hi

Lozansky

I had a midterm on E&M today

Semiclassical

neat

how'd it go?

Lozansky

18:41

Think I nailed it

Which is never a good sign

Semiclassical

lol

Lozansky

First question was on linear dielectrics, so I was kinda prepared for that :P

Second question was a bit more challenging

Given a sphere of radius $a$ centered at origin and with constant polarization $\mathbf{P} = P_0 \hat{z}$ and angular velocity $\mathbf{\omega} = \omega \hat{z}$, find the magnetic field at the north pole on the sphere

Semiclassical

huh

user280247

@Semiclassical I'm not saying it doesnt exist, I'm saying I couldn't find the argument over the web

user280247

@EmilioPisanty didn't get what u mean

Semiclassical

18:49

@Lozansky yeah, that seems pretty tough

getting the electric field is tedious enough

user280247

@EmilioPisanty I've done that for simple shapes...

Semiclassical

I guess the idea is that you'll have bound surface charge on the sphere and this rotating charge gives rise to a magnetic field

finding the former isn't bad

but i'm not sure how hard the latter is

Lozansky

You are on the right track indeed

enumaris

I can't remember really anything about polarizations

lol

Lozansky

(Or, at least the same track I ventured on)

Semiclassical

18:53

you'll have a surface current on the sphere, to be specific

Novalium Company

Hello everybuddy!

enumaris

Probably you can make some arguments from axis-symmetry and draw an ampere loop and be done with it :P

Novalium Company

I was wondering if a tesla coil (with alternating current throught it) does not kill a human because the voltage is very high, but the current is very low? (conservation of energy, tranformers...). This tesla coil will produce an alternating magnetic field around it and when we bring a light bulb for example closer to the tesla coil, the alternating magnetic field will induce alternating current in the light bulb and that's why it lights up. Is this true?

Lozansky

@enumaris Ehh... doubt it

Semiclassical

eh. not sure you'll be able to get away with Ampere's law here

Novalium Company

18:56

Also in the end, in order for a tesla coil to not kill a human, the input supply power mustn't be high enough to kill a human, and from there, whether you'll use a transformer to level up the voltage and lower the current, it doesn't matter?

coniferous_smellerULPBG-W8ZgjR

elo all

Semiclassical

Biot-Savart will definitely work, and is probably necessary

Lozansky

The local velocity is proportional to the surface current

@Semiclassical That's how I did it

enumaris

well, no clue, but it feels like it could be one of those "haha, there's this trick everybody missed" kind of problems

maybe I'm just paranoid...

Lozansky

More like: there's this trick - but then there's this integral.. :(

danielunderwood

19:00

Ew Biot-Savart

Semiclassical

yeah

Novalium Company

@Semiclassical Could you please if you have some time help me with my understanding of tesla coil?

Semiclassical

no

Novalium Company

@Semiclassical Ok, sorry.

Anonymous

@NovaliumCompany Sounds okay

Anonymous

19:05

As long as the load current is low, it shouldn't matter (shouldn't be sufficient to kill)

Novalium Company

@Blue Oh, thanks. I asked my quesitons a bit up in the chat.

Anonymous

The power is the constant quantity, in transformers. You're right about that

enumaris

@NovaliumCompany en.wikipedia.org/wiki/Tesla_coil#Health_hazards seems like tesla coils can kill someone

quote: "Even low power Tesla coils could exceed these limits, and it is generally impossible to determine the threshold current where bodily injury begins. Being struck by arcs from a high power (> 1000 watt) Tesla coil is likely to be fatal."

Anonymous

"high power" being the keyword

enumaris

I think the take away is don't mess with it if you don't understand it lol

coniferous_smellerULPBG-W8ZgjR

19:09

@Blue I got 100 answers + qeustions though i know you dont care an iota about numbers ^^

but for me it's like it means my life's over i can't post any more question nor answer

Novalium Company

Got it, thanks. So is there any difference between high voltage/low current strucking the body or low voltage/high current stucking? I mean, if they are the same power (P - const) is there any difference to what damage can be done? (Will a person feell more pain in high voltage/low current or low voltage/high current)?

Emilio Pisanty

@Blue @JohnRennie @ACuriousMind bam.

single-self-answer HNQ

Anonymous

@NovaliumCompany $P_{supplied} = V\times I$ and $V = IR$. If power is same and the body resistance is same, then current and voltage automatically get fixed

Emilio Pisanty

and all it needed was a little bit of clickbait in the title

Anonymous

@EmilioPisanty Hah, so you cracked the HNQ secret

Anonymous

19:12

:D

Semiclassical

I think the complication will be that the human body probably doesn't act like an ideal resistor

Emilio Pisanty

@Blue on occasion, at least

I still think it's a scourge and it should be heavily reformed, if not done away with

but since it's there, I'm not above gaming a couple of its aspects to get material I think is important in front of a wider audience

coniferous_smellerULPBG-W8ZgjR

wuut? you got a nobel prize @EmilioPisanty???

Anonymous

@Semiclassical That's true. Once the initial resistance is overcome by the current, it rapidly reduces

coniferous_smellerULPBG-W8ZgjR

or did i misunderstand

Emilio Pisanty

19:14

@coniferous_smellerULPBG-W8ZgjR what?

no

not at all

Anonymous

@coniferous_smellerULPBG-W8ZgjR Watch the news :P

Emilio Pisanty

I wrote a Q&A about a Nobel prize

coniferous_smellerULPBG-W8ZgjR

i dont have tv

Semiclassical

wow we've got a nobel laureate woooow

coniferous_smellerULPBG-W8ZgjR

ahh

Emilio Pisanty

19:15

6

Q: What is Chirped Pulse Amplification, and why is it important enough to warrant a Nobel Prize?

The 2018 Nobel Prize in Physics has just been announced, with half going to Arthur Ashkin for his work on optical tweezers and half going to Gérard Mourou and Donna Strickland for developing a technique called "Chirped Pulse Amplification". While normally the Wikipedia page is a reasonable place ...

coniferous_smellerULPBG-W8ZgjR

well congrats either way @EmilioPisanty ^^

Anonymous

@Semiclassical lol

coniferous_smellerULPBG-W8ZgjR

upvoted from my part. just read the qestion

enumaris

@Semiclassical who's the "we" in that statement? o.o

Anonymous

Photonics/Optics should remain a hot area for the nobel prize winners in the coming days too

Semiclassical

19:16

physics stack exchange. (i was not being serious)

Anonymous

It's one of the most useful areas

Anonymous

and has dem....APPLICATIONS

Emilio Pisanty

@enumaris the wider community of physicists

we actually got three

Anonymous

Any fields medalists on PSE?

Emilio Pisanty

@Blue as in, they have accounts?

enumaris

19:18

@EmilioPisanty But there's a nobel prize specifically for physicists...so every year we get 1...(or 2 or 3)...

Anonymous

@EmilioPisanty Yes (and a couple answers maybe)

Anonymous

Who are the 3 people (Nobel prize winners on PSE) tho? I know only T Hooft

Emilio Pisanty

@enumaris I stand by the truth of my assertion, though I make no claims of non-banality

enumaris

hmmm...

Emilio Pisanty

@Blue I imagine you've seen this guy

Terry Tao, Los Angeles

56k 17 241 333

nt.number-theory big-list analytic-number-theory ca.classical-analysis-and-odes soft-question co.combinatorics fourier-analysis

Novalium Company

19:21

@Blue So in the end, does it matter if high voltage/low current hits the body or low voltage/high current, where P is constant? Will the person still feel the same pain?

Emilio Pisanty

but he doesn't have a PSE account

Anonymous

@EmilioPisanty Yeah. Everyone knows him on MO :P

Anonymous

There actually may be more fields medalists on MO using anonymous accounts

Anonymous

And likewise for PSE (as for nobel)

Anonymous

[avoiding the limelight]

Emilio Pisanty

19:23

bam! rep-cap.

::grin::

Anonymous

@NovaliumCompany The current passing through your body is ultimately what hurts. SemiC mentioned an important point i.e. if the initial voltage is not high enough, then it might not even overcome your body resistance (the resistance of the body only goes down once the current gets a low resistance path to flow)

Anonymous

So you should probably use a high voltage if you want to kill

Novalium Company

@Blue Ha, thanks :D

I'mma go to bed now, I have school tommorow, have to wake up at 6:15AM :( Thanks everyone for the help.

Anonymous

6:15 am school ;_;

Anonymous

I will shed some tears on your behalf

coniferous_smellerULPBG-W8ZgjR

19:29

@Semiclassical We have Gerard T Hooft

Novalium Company

@Blue I barely even know who or where I am when I wake up :D My legs just start doing their thing and taking me to my Six Cruel Hours Of Our Life.

Anonymous

lol

Emilio Pisanty

19:39

hmmmmmmm.

@Qmechanic, I notice you have a bunch of Announcer badges

do you regularly post Share links outside of SE?

or is this answer inaccurate?

9

A: Do shared link to ANSWERS also count for Announcer, Booster, Publicist badges?

No. These badges are specifically for linking to questions like the description says. The link must be clicked from outside the network in order for it to count for this purpose.

ACuriousMind

19:55

@EmilioPisanty I have 58 announcer badges and I know that I haven't posted 58 links with the bit identifying me as the origin of the link outside of it.

Emilio Pisanty

@ACuriousMind hmmmm.

ACuriousMind

13

A: Announcer badge spike

This is caused by the switch to HTTPS. Stack Overflow changed to HTTPS on May 22, and guess what happened. Here are the awarded Announcer badges on Stack Overflow: This has a technical reason. Visitors coming from within the Stack Exchange network didn't count towards the badge as they were f...

^https is the culprit

or maybe not, see the comments, but Announcer is definitely not working exactly as it should

Emilio Pisanty

@ACuriousMind huh

that reminds me, though

Emilio Pisanty

20:11

0

Q: The mainstream-physics close reason has a link to meta with the user ID of a moderator. Can it be removed?

Basically, what it says on the tin. The mainstream-physics closure dialog, which is accessible on the closure dialog as well as on closed questions (example), contains a link to the mainstream-physics site policy which contains the user ID of one of the elected site moderators, i.e. the same link...

discussion close-reasons non-mainstream-physics hyperlinks

Physics Meta

20:22

0

Q: The mainstream-physics close reason has a link to meta with the user ID of a moderator. Can it be removed?

Basically, what it says on the tin. The mainstream-physics closure dialog, which is accessible on the closure dialog as well as on closed questions (example), contains a link to the mainstream-physics site policy which contains the user ID of one of the elected site moderators, i.e. the same link...

discussion close-reasons non-mainstream-physics hyperlinks

Semiclassical

lol

enumaris

hmmm

Qmechanic

@EmilioPisanty : No, rarely.

danielunderwood

20:38

Does every meta post automatically get linked in here?

enumaris

it would appear so

reading Woit's blog and got to IUT again...fun times...

@danielunderwood I took a deeper look, and gensim is explicitly only for topic modeling using LSI or LDA or other such models.

danielunderwood

IUT?

enumaris

Inter-universal teichmuller theory

that 1000 page thing the Mochizuki published in 2012

Leaky Nun

@enumaris ...

enumaris

sup

danielunderwood

20:49

That sounds a bit out there

Leaky Nun

inf

enumaris

It seems Mochizuki is quite protective of his theory

Semiclassical

notably, woit didn't post anything about the whole Atiyah thing (and indicated that his doing so was deliberate)

danielunderwood

And aha! I got you to do my work for me. I was actually planning on figuring out something for topic modeling though, so that sounds like a good start

Semiclassical

which in retrospect was a respectable choice

enumaris

20:50

going so far as to accuse SS as such: "on the part of SS, of their profound ignorance of the elementary theory of heights, at the advanced undergraduate/beginning graduate level."

@danielunderwood lol, yeah, I mean the basics of topic modeling is very similar to low-rank-matrix factorization used for collaborative filtering

or PCA

Gensim just makes those algorithms more accessible

It's unsupervised so you don't run into the same problems you do in sentiment analysis like inter-annotator agreement

Leaky Nun

@Semiclassical who is woit?

Semiclassical

peter woit

mostly notable as a writer/blogger

Curious Fish

u woit m8

enumaris

clever

Curious Fish

this website is great

so

y'all wanna talk abt something

barlop

21:10

What is my (real/physics) mass, if my weighing scale says I weigh 70kg?

danielunderwood

@enumaris String stheory?

And I should probably look more into PCA. I hear data people talk about it a fair bit. I think it's a lot like SVD?

Emilio Pisanty

@Qmechanic so the Announcer badge is broken then?

oh well

Student404Mus

21:41

hi

if Higgs boson is the responsible factor which generate masses to other bosons, after the SSB where is goes ?

vzn

21:55

speaking of Tao just discovered from offhand blog comment hes lately working on a numerical/ number crunching Riemann attack terrytao.wordpress.com/2018/09/06/… michaelnielsen.org/polymath1/… reddit.com/r/math/comments/9jy1ai/…

enumaris

@danielunderwood the two are highly related

danielunderwood

I suppose PCA may work if there aren't singular values or the arrays aren't square? (I think you can actually get singular values of non-square arrays, but I haven't messed with that)

enumaris

@danielunderwood in particular, SVD decomposes the original data matrix into 3 matrices - one of which will correspond to the matrix of eigen-vectors of the correlation matrix that PCA decomposes.

Singular values decompose the data matrix X

PCA decomposes $X^T X/(n-1)$

I mean, for normal matrices, the singular values are actually just the absolute values of the eigenvalues so that should tell you how similar they are

danielunderwood

Our mechanics professor mentioned PCA when we were doing inertia tensor stuff and I wrote down to look into PCA...that was over 2 years ago I think

enumaris

heh, for the inertia tensor, I suppose PCA will find the principle axes

danielunderwood

22:09

Well we did SVD stuff and he mentioned that PCA was an extension of it used in data analysis or something to that extent

I suppose in physics we mainly just deal with normal matrices though

enumaris

yeah

danielunderwood

Although I don't really think "hmmm is this a normal matrix?" every time I come across a physics problem with a matrix

enumaris

Hermitean matrices are Normal so

so are symmetric matrices

erm

sorry I mean symmetric real matrices

Semiclassical

it's not quite true that the only matrices we care about in physics are hermitian, but it's definitely true a lot

enumaris

unitary matrices as well...

vzn

22:13

@Semiclassical it seems strange woit would write so much about ABC but zilch )( about Atiyah. 2 cases seem similar to me.

enumaris

Unitary, Hermitean, skew (anti) Hermitean, real orthogonal, real symmetric, real skew (anti) symmetric...

Semiclassical

@enumaris "unitary matrices" are probably the big exception to "physics only cares about hermitian matrices"

enumaris

yeah, they are normal too though

Semiclassical

since $e^{-itH}$ is unitary but very much not Hermitian

enumaris

so...that's a pretty wide class :P

Semiclassical

22:15

Yeah

enumaris

mathematicians would disown you if you said "singular values are just the absolute values of eigenvalues" though

Semiclassical

lol

danielunderwood

22:31

Well that's better than what I thought. I thought singular values were exactly the eigenvalues if you had a square matrix

But I only had 1 class that mentioned singular values as a thing and we only did square matrices in that class anyway

enumaris

I didn't have any classes where singular values were a thing

just eigenvalues

looks like gensim will be actually pretty useful in the future...

maybe even if I don't do topic modeling...just for some of its preprocessing features

danielunderwood

I think my numerical analysis professor briefly mentioned it while doing SVD. I don't know why we didn't actually do anything with non-square matrices in that class. We also went over QR factorization, which I haven't used but feel like it should be useful somewhere

I'll probably mess around with gensim a bit once I get back around to messing with tweets

enumaris

it can do bag of words pretty easy

and then bag of words to TFIDF

It might have saved me a bit of time from writing my own TFIDF algorithm lol

but I mean it was only a few lines long so maybe not that much...

plus my version has a tf-idf threshold hmmm

Transcript for