| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Comments:
<0> jdh41: I want to understand the stuff myself!
<1> Numerical recipes is from memory lots of code snippets which people can copy in for fun and profit
<2> Aldaron: I read it, but is essentially on how to make some math operations in C++
<0> J_B: oh, so it doesn't really even comment on the math involved? =P
<2> Aldaron: didn't say that
<2> it's not a bad book in the end
<3> if you want code that is
<2> just nothing that will solve all your problems ;)
<3> and a can say a fair few do
<1> Aldaron: Do you know any academics in that area who you can ask for a recommendation?
<0> J_B, do you know better books? If I want to focus on the math, but want it such that I can turn it into code myself! (just a formula and the name of a method obviously don't do it.. yet :(
<4> hello
<0> jdh41: now, that's a good suggestion, I'm working with them ;)
<4> do you know how to proof that supp(f*g) is subset of supp(f) + supp(g)? (with f*g (x) = int f(x-y)g(y)dy and A+B = {a+b: a in A, b in B}
<0> Maybe I should browse through a library too
<1> yeah, reading existing code good
<1> I imagine the libraries probably come with papers if you've got access to online journals
<3> /me's antivirus just crashed and he can't kill it cause it's protected... the joys of trusted computing
<3> /mefoo
<3> wierd
<0> jdh41: I do have access, and our library has a vast amount of journals too. And books. I can definitely find all the differential equations I could hope for ;). Turning those into a program is what I can't do :P
<1> There shuold be papers on how to do that I'd have though.
<3> J_B: damn, the NRiC PDFs need some sort of esoteric plugin install to open, definitely not my taste
<3> thanks anyway
<0> jdh41: Any ideas on how to find them? :P
<3> Aldaron: and why not?
<3> you could even base stuff off the GSL or similar
<1> google scholar/something similar for the name of packages others have written in related areas will probably turn up the correct journals to be looking for
<0> Jafet, I lack.. something! ;)
<5> what is sum(1/x^2) worth?
<5> for x all natural integers
<5> except 0 :)
<0> Might be that I lack the english to know which functions I need to use. I don't see a "simulate an atmosphere at given resolution" function in GSL either :P
<3> retybok: about $2.99? (:
<3> 6/pi i think
<3> Aldaron: you have the math, libraries gives low-level stuff, surely you can craft whatever's in between
<5> Jafet: thanks
<5> Jafet: actually, it's Pi^2 / 6 :)
<5> according to wikipedia
<3> whoops
<3> well there you have it
<0> Jafet, then I'm not sure I have the math well enough :P. I've coded a simulator for 2-d heat diffusion, but I did that with lots of help.. And something that includes m*** transfer, velocities et cetera at least sounds much scarier! ;)
<3> Aldaron: maybe it's the language you're using
<3> like, i wouldn't code a web script in fortran
<3> C++ isn't what i'd call easy either
<0> Jafet: might be. I know I should learn fortran better, and f90 too for that matter ;). The diffusion progam I wrote was in f77
<2> Jafet: NRiC PDFs opened without any problems to me :|
<3> it needs some sort of "fileopen" plugin which requires ActiveX to install and kinda, um, gives me the creeps you know
<2> Jafet: strange, when I downloaded it (it was a couple of years ago) nothing of this happened :|
<0> All right, any ideas about a book that comprehensively covers fluid modeling down to the code? ;)
<3> the website said they updated the ebooks so i guess it all boils down to bad timing
<3> Aldaron: good luck (;
<0> Jafet, thx X-). I'll first get the free numerical recipes thing at least
<2> Aldaron: if you know fortran be sure to search for the fortran numerical recipe ;)
<0> J_B: I was thinking of the fortran version. I know enough fortran to understand the code, and I get to learn fortran too.. :)
<2> Aldaron: I think that in the end you are searching for some engineering book more than some math book (am I wrong?)
<0> J_B: I guess you're right. I do need the math too, but I think I already have those books
<0> Then again, there's not much left to learn in pure programming books
<2> what do you need to model in a fluid?
<0> J_B: Everything! Well, at least temperature, salinity, transport (all of it), atmospheric forcing..
<0> And when I can do that, add ice cover and ecosystem modeling and.. ;)
<6> when it finally comes down to it, it's always about our precious math ;-)
<0> always :)
<2> "everything" means nothing :-P Do you need to study some wird liquid crystal charge transport properties or to study osmosis of water into cells?
<3> Aldaron: um... why? i don't think even the big government labs do it
<2> *weird
<3> you could model individual particles or flows CA-style and leave it as that
<0> J_B: okay, let's clarify :). I'm interested in modeling (1) oceans, (2) atmosphere
<3> Aldaron: for 1) animation 2) research 3) entertainment
<0> Jafet: 3 and 2, in that order
<0> Jafet: I work for a (big?) government lab, as a research ***istant :P
<0> ..And I would like to understand the models I'm working so much with
<3> well, you won't find any research effort trying to squeeze in more variables than it can use
<2> Aldaron: doesn't your big boss know all possible books on the subject?
<3> a bureacratic PHB? unlikely i say
<0> J_B: uhm.. My big boss wrote the model I'm working with most, and I tell you, he doesn't know much about programming at all ;)
<0> I'm sure he could teach me, if he got the time to do it
<7> it's so painful to see people not knowing how to code properly, code
<0> jengelh: Tell me about it. Been cleaning up a lot of stuff around here
<2> Aldaron: you can read his papers on the model and undersatnd the physics first. Than find out which operation you need actually to perform and search good books that teach you how to implement that and finally optimize the code ;)
<0> How about writing a fortran program to spew out ten shell scripts every day, just because he couldn't use variables in the shell scripts in the first place? :P
<7> Aldaron: Students and teachers who just program to implement something (simple vector graphics, opengl, you name it) do so poorly
<0> J_B: That's a good idea too - walk to him and ask what papers he's written on the subject and where to get them ^_^
<7> demo programs not included.
<0> J_B: Although the papers I've seen this far don't really go into the implementation part at all. "basic formulas blahblah data blah conclusions nice pictures!"
<2> jengelh: 99% of time it is enough (my code are never optimized, anyhow I use most of them not more than a dozen time)
<7> well perhaps I am just too used to good-and-clean code, as Linux kernel code review dictates :)
<2> jengelh: not everyone can be a coder ;)
<7> so if You can't do it, let someone else
<7> You make the theory, let someone else do the coding
<7> that's what science is for!
<0> J_B, not optimized is okay. Something that doesn't compile anywhere else or run without the exact funny tweaks that the writer had is BAD.
<7> for those who like to eat peanut butter, what do you eat it with?
<2> jengelh: you means you are willing to code a light transport velocity into Mie systems programs for me? Thank you so much!!! :D
<8> if f(x) = sum_{n=1}^{\infty} a_n phi_n(x) ; will then f'(x) = sum a_n phi'_n(x) ?
<7> J_B: I said clean programming. Not extending it beyond purpose.
<7> Though we might talk about it ;-)
<3> jengelh: so no porn sites? *cries*
<7> what does porn have to do with it
<2> jengelh: my code is already working :P It just take 10 times the time a good code would take (the fact that I made it with Mathematica may have something to do with it :") )
<7> I was thinking along the lines of regular coding, C, C++, Java, Perl, etc.
<7> Math languages are a different matter (there is not much you can actually do wrong)
<8> so what about my problem with the series (to get back to math)
<8> i know f(x) = ... is satisfying the M test and hence is both absolute and uniformly convergent
<8> i know f is twice differentiable
<8> what do i know about the series expansion? :)
<2> holst: differentiation is a linear operator so I think it's ok to transfer it inside the sum (If it is converging)
<9> DEATH TO ****IN ALGEBRA
<9> Another ****in D, this is bullcrap
<9> !
<8> hehe
<3> woah woah kid can we sit down and talk about this
<9> Bull**** enough talking
<9> I had a D Before
<9> I retook that ****in course to improve the grade
<10> is this a declaration of war?
<3> calm down dear it's only a chatroom
<9> by those collection of proffesors in this course ARE NAZIS
<9> NAZIS
<9> they rape you when it comes to reducing points
<9> its bull****!
<9> hey ! he forgot a definition! BOOM -10 points
<7> hm reminds me Godwin's law
<11> hey come on, keep the language clean please
<7> hm reminds me of Godwin's law
<8> never seen math induce such temper
<3> should we send this guy for counselling or just kline him
<7> perhaps both
<9> god damn bull****
<8> i thought that was for coding only and system administration especially. well anything, related to computers
<9> im gonna play some wow
<7> wow = teh lame
<3> ...what a r-tard
<8> thats his problem: too much WoW and to little algebra
<8> :)
<9> yea but there people accept me for what i am !
<9> Algebra is ****
<7> we don't accept you right now
<9> i hate it
<10> bad at math?
<9> calculus is fun
<2> Copter>yea but there people accept me for what i am ! <-- an orc? :P
<9> Shadow undead priest rlly
<8> a troll!
<11> uh, you need algebra to do any serious calculus
<10> you should do some algebra instances
<3> solve X = raging math-hating kid who likes WoW... see? it's that easy
<12> Who did the UKMT today?
<0> Would someone happen to have the Numerical Recipes PDFs without the FileOpen screw-everything plugin?
<7> use xpdf?
Return to
#math
or
Go to some related
logs:
the specified module could not be found websvn
ubuntu kill xmms
ubuntu mount privelages
gdm suse10.1
#physics
thang online wine
spikebike
#math
ml 2010 yast crashes
livna Metadata file does not match checksum
|
|