Large Hardon Collider

[Kaylia]

Uh?

When I attended Krauss' conference last year at Mcgill university,

Not talking about the video.

[Eliseos]

my lab, nothing special or excitang

Even that giant red button looks new

[Kaylia]

I would totally press it. Accidently.


That labs almost look too recent. I wouldn't dare playing football with the broken material in this room.

[Penthesilea]

Eh, it cost ~ a mil but is a piece of shit so... yeah we've horsed around in there a bunch.

EDIT: sorry was mistaken on what specific talk you were referencing. In any case, scientists can provide input into ethics discussions and if that's what you meant to say you're wrong. If you meant that ethics can be proven by science, ok I'll give ya that.

Thanks guy who posted the video, I found it interesting and I have little to no experience with physics/cosmetology

[Eliseos]

What is it supposed to do?

[Penthesilea]

We do uHTS or high through put screening. It's just one of the robot platforms that run the experiments that I develop. Essentially we've got poly propylene plates with various well densities that contain our different small molecules. Each plate is run against a target to see if there's any difference in signal, signifying that the compound could be a novel lead for a drug.

It's basic biochem stuff, nothing major on my part.

[Eliseos]

A basic experiment for a chemist that costs a mil?

Woozie posted this a few pages ago but it's relevant again

http://www.smbc-comics.com/comics/20081008.gif

[Penthesilea]

Well yes and no. It's about making it worth while to screen 2.3million different compounds. If we were to run those experiments by hand in 96 or 384 well formats it'd take forever. As is it takes about 9 days to screen the entire library against a single target.

[Tristam]

Anyone here used Monte carlo methods before to simulate system? Just curious since it's what I'm working on right now.

Yes and no, I used bootstrap monte carlo simulations to generate confidence intervals for biochemical kinetics data. But the code was entirely written by someone else and I barely have a wiki level understanding of the statistics. Before that I did some modeling exercises in a graduate level computer programming class that used some monte carlo simulations iirc, I discovered I write code about as well as I fly rocket ships and I haven't touched a computer for programming purposes in over a decade.

[Kaylia]

Another help request for my homeworks! (2 in 2 weeks...shame on me)

Basically, we have to demonstrate lagrange multiplier in details (which is something I used quite a lot, but never understood how it works). I found a few demonstrations on the net (wiki has one), but I would lie if I said they made sense to me right now. Anyone know where I could find a slightly more obvious demonstration before I head to the library for a better answers (e-books or website would be the best).

Yes and no, I used bootstrap monte carlo simulations to generate confidence intervals for biochemical kinetics data. But the code was entirely written by someone else and I barely have a wiki level understanding of the statistics. Before that I did some modeling exercises in a graduate level computer programming class that used some monte carlo simulations iirc, I discovered I write code about as well as I fly rocket ships and I haven't touched a computer for programming purposes in over a decade.

I haven't done much programming recently, but since my degree involve a lot of simulation, I need to get back into it...very soon. I've been studying differents methods for two weeks, but I still have no idea how I'm going to simulate the problem I'm dealing with (hydrogen adsorption mostly). I was simply curious to hear people experience with it in the past.

[Ferion]

Been a long time lurker of this thread, don't know too much about Physics, but I am into math so I find a lot of what you guys talk about very interesting. Kaylia, I'm not really sure if you mean you want a proof for Lagrange Multipliers or an explanation about why/how it works. Either way, I found these two links to be very helpful although you may have found them already.

http://www.slimy.com/~steuard/teachi.../Lagrange.html
http://xbeams.chem.yale.edu/~batista...grangeMult.pdf

The second contains a short but pretty clear proof.

[Kaylia]

I saw the first link yesterday, but I don't see any demonstration (just some explanation). The second link was exactly what I was looking for, so thanks a lot, you just saved me one trip to school!.


[edit]
one part of the sentence was missing.

[Ferion]

No problem. I don't know if this will make sense for Physics, but I'm kind of curious, what mathematics is most heavily emphasized to know? I know this is very vague and probably depends strongly on what you study, but for example, with Economics almost every adcom for graduate school mostly cares about proof based math classes.

[Kaylia]

I'm tempted to answer "all of them". Like you said, different physics topics will favor differents mathematics, but in the end, there isn't many fields that are not used in physics.

Algebra and vector calculus are definitively useful in every field of physics. You need good understanding of partial derivative, differential equation, lebesgue integral and various thing like this to be able to solve most problems. How much you need to understand depend of the class and fields, but one thing is sure, you will deal with them, and a lot. If you go toward material sciences and physics statistics, you will see a great deal of statistics (duh) like distributions. If you go toward QM instead, linear algebra usually become dominant, and you will deal with matrices, hilbert spaces, operator. Basic geometry will be used everywhere (even if it's just for an approximation), but differential geometry will play a key roles in relativity. Electromagnatism and fluid dynamics consist of vectorial analysis (curl, gradiant, differential equation).

While none of my class dealt with analysis directly, I do think it's important to grasp the basics stuff to understand the maths you're using.

People might disagree with this, but it's the overall impression I got from my major. If you are looking for even more specific topics, you will usually end up dealing with something difference since you combine more than a single field (for example, studying a material in extreme condition often involve quantum mechanics and physics statistics..and all the mathematics they uses)

[Mizango]

Nice to meet you Ferion :D Post moar.

[Eliseos]

You're going to also want to be really really good at simple math. Today on an exam I decided that (1/3)*(1/3)*(1/3) is (1/9) instead of the correct (1/27). * != +

I think that is the only part I got wrong, but it is still infuriating that as you take more and more advanced classes in math/science/engineering, your ability to do simple math severely diminishes. I probably get more points taken off on homework/exams for stupid things like losing minus signs or not being able to add properly, than I do for not getting the material portion correctly.

Edit: I just realized that (1/3) + (1/3) + (1/3) isn't even (1/9), it's just 1

[Kaylia]

You're going to also want to be really really good at simple math

I can't tell you how many time I screwed up matrice filled with 0 and 1, or something similar. It makes me sad everytime.

and lol at the edit. I didn't even catch it the first time I read it, so don't feel too bad!

[Woozie]

As Kaylia says, you'll need a huge variety of math. I have no idea where you are in math or physics. If you're an undergrad, you'll want calc1->calc3, diffEQ, and linear algebra at minimum. After calc 3 (or after calc 1 but it will be more of a challenge), you could get a book on engineering mathematics or mathematical methods for the physical sciences (Mary Boas writes a really good book for this). Those books will give you the extra bits of math that you probably wont pick up in a math course but will need in physics (e.g. sturm liouville equations, fourier analysis, laplace/poisson equations, calculus of variations, etc).

Whether or not you need advanced mathematics and which ones you'll need depends entirely upon what you choose to study.

Also, this:

Nice to meet you Ferion :D Post moar.

Edit: Yeah, and what Eli and Kaylia said...try not to forget the simple math. Last week I argued with my professor about why subtracting a quantity from itself gives you zero. We had iy-iy = 0 and I couldn't see why that had to be true. To be fair, that class is at 7:45 in the morning, but still....

[Ferion]

Thanks for the explanation Kaylia, that's basically what I figured. I'm kind of surprised that Physics students aren't required to take some sort of Real Analysis course, it's really tedious at times but pretty damn useful to know.

Nice to meet you Ferion Post moar.

Haha nice to meet you guys too.

As Kaylia says, you'll need a huge variety of math. I have no idea where you are in math or physics. If you're an undergrad, you'll want calc1->calc3, diffEQ, and linear algebra at minimum. After calc 3 (or after calc 1 but it will be more of a challenge), you could get a book on engineering mathematics or mathematical methods for the physical sciences (Mary Boas writes a really good book for this). Those books will give you the extra bits of math that you probably wont pick up in a math course but will need in physics (e.g. sturm liouville equations, fourier analysis, laplace/poisson equations, calculus of variations, etc).

I'm a undergrad in my senior year majoring in Math and Econ. I don't really plan on getting into physics in college, just find a lot of the concepts very interesting. I was just curious what kind of math you guys need to primarily use cause a lot of the stuff I see in this thread tends to be integration problems. But I do appreciate the advice, I may end up picking up one of my brother's engineering books if I have some free time. It's kind of strange to me how different but similar the math you guys use compared to what I study. For example, Brownian Motion and stochastic calculus is used in option pricing and stock market analysis.

[Woozie]

Physicists don't care about being rigorous in their math (which is the thing I hate most about physics).

Though an understanding of real and functional analysis makes a HUGE difference when trying to learn quantum theory on the advanced level.