|
Bloody posted:what is the least painful lang to janitor text files OR a sqlite database with and fart out plots python by a mile. bokeh looks promising and there's a guy from the company posting in the coc python thread http://bokeh.pydata.org
|
# ? Nov 19, 2014 05:17 |
|
|
# ? Apr 17, 2024 23:15 |
|
Bloody posted:what is the least painful lang to janitor text files OR a sqlite database with and fart out plots FORTRAN
|
# ? Nov 19, 2014 05:31 |
|
I'm using mathematica right now
|
# ? Nov 19, 2014 05:48 |
|
in matlab 2013 or something (like a thousand american dollars) with the miltiprocessing toolbox (like another thousand american dollars), you can use cuda to do some poo poo. but with python you can use pycuda and just plop cuda function into your code as triple quote blobs. don't use matlab, op
|
# ? Nov 19, 2014 05:49 |
|
bump_fn posted:maple is poo poo for babies aka canadians we were taught maple in school. it was ok. good enough for just gettin poo poo done. ugly as sin though
|
# ? Nov 19, 2014 06:41 |
|
in other news I learned about seaborn today im never using vanilla matplotlib again
|
# ? Nov 19, 2014 06:50 |
|
agh thats the answer i was afraid of
|
# ? Nov 19, 2014 06:50 |
|
one time i wrote a surface automorphism function in maple to watch wtf happened to rational numbers. then my advisor was like, "yeah but what happens to rational polynomials?" and then i tossed a rational polynomial into the automorphism function and it just worked there was even a call to gcd or something and apparently the gcd algorithm just worked for polynomials instead of integers. p cool
|
# ? Nov 19, 2014 07:04 |
|
Bloody posted:agh thats the answer i was afraid of Yeah there just aren't that many good in-language plotting tools. MATLAB sucks. matplotlib is a small improvement but is mostly bad in the same ways as MATLAB One really good in-language plotting library is ROOT, which is callable from C++ and Python, but it has a huge learning curve and has its own problems (like ugly default settings). I don't recommend diving in unless you're already part of a group that uses it
|
# ? Nov 19, 2014 17:51 |
|
maple is very good and nice to use, although i have never been able to get its stupid partial fraction tool to work for any numbers other than their examples.
|
# ? Nov 19, 2014 18:09 |
|
Mathhole posted:it just worked its pretty neat how you can gently caress up your math something fierce and it might chug for a bit but will almost always pop out some kind of an answer (involving ten pages of integrals and a branch of math you've never even heard of) makes getting a nice concise result v. satisfying
|
# ? Nov 19, 2014 19:30 |
|
Hammerite posted:but then unless I've misunderstood you, that's not a matrix as such, just a bunch of data in a table that you've decided to call a matrix. it's a glorified array. if it wouldn't be meaningful to do matrix math operations on it like, i dunno, finding its eigenvectors or w/e then it wouldn't be meaningful to do matrix multiplication with it. if, conversely, it would be meaningful to do matrix multiplication with it then i'd imagine it would not be meaningful to do elementwise multiplication. you're right in a mathematical sense, however electrical (signal processing) engineers give no fucks and guess who uses matlab
|
# ? Nov 19, 2014 19:44 |
|
i want to write steven wolfram a strongly worded letter about his riemann theta function conventions
|
# ? Nov 19, 2014 19:51 |
|
Mathhole posted:one time i wrote a surface automorphism function in maple to watch wtf happened to rational numbers. then my advisor was like, "yeah but what happens to rational polynomials?" that aspect to it is neat but it sucks to write functions (sorry, procedures) in
|
# ? Nov 19, 2014 20:01 |
|
hobbesmaster posted:you're right in a mathematical sense, however electrical (signal processing) engineers give no fucks and guess who uses matlab He's right that a matrix is just a special type of Array, and you should just always use Arrays because you can still apply matrix operations to them, just with a function instead of a symbol
|
# ? Nov 19, 2014 20:28 |
|
ok how do i set up a python to be like a better matlab and then how do i use it like a better matlab
|
# ? Nov 19, 2014 21:38 |
|
why are there two major versions of python which one do i use wtf
|
# ? Nov 19, 2014 21:41 |
|
use 2.7, install numpy, scipy, matplotlib, and as many of the "scikit" packages as you need for your stuff (scikit-learn and scikit-image) are the big ones also get ipython, and if you're going to do interactive work always always always work in the ipython shell or notebook
|
# ? Nov 19, 2014 21:50 |
|
you should probably ask python questions in the python thread, this thread is for making GBS threads on MATLAB Like who the gently caress thought it was a good idea to return an int8 if you add an int32 to an int8 And why even bother having an int64 if you don't define any operators for that datatype. The MATLAB int64 and uint64 have existed since the early 2000s, but operators for them weren't added until 2012. Hope you guys don't want your math toolkit to let you use large numbers QuarkJets fucked around with this message at 22:04 on Nov 19, 2014 |
# ? Nov 19, 2014 22:00 |
|
Do that stuff fritz said and also use pandas. Jetbrain's pycharm has a good free version if you want to get fancy and ipython notebook isn't cutting it. Also a 3rd party package manager (conda, canopy) can be helpful, especially if you're on windows.
|
# ? Nov 19, 2014 22:02 |
|
Spime Wrangler posted:Do that stuff fritz said and also use pandas. Jetbrain's pycharm has a good free version if you want to get fancy and ipython notebook isn't cutting it. the first step for ipython was to install anaconda so im doin that
|
# ? Nov 19, 2014 22:04 |
|
Regardless of whether you use Linux or Windows, I'd recommend just downloading anaconda. It makes management way easier, especially if you're on Windows but it's still good on Linux e: Good advice happening here
|
# ? Nov 19, 2014 22:05 |
|
quote:Traceback (most recent call last): cool off to a great start
|
# ? Nov 19, 2014 22:07 |
|
QuarkJets posted:return an int8 if you add an int32 to an int8 lol QuarkJets posted:And why even bother having an int64 if you don't define any operators for that datatype. The MATLAB int64 and uint64 have existed since the early 2000s, but operators for them weren't added until 2012. Hope you guys don't want your math toolkit to let you use large numbers lmao
|
# ? Nov 19, 2014 22:25 |
|
Bloody posted:cool off to a great start This has still got to be better than wasting any time loving around with Cygwin.
|
# ? Nov 19, 2014 22:27 |
|
update: i fixed whatever the gently caress that was before and now i get to joust with getting python poo poo to work behind a proxy
|
# ? Nov 19, 2014 22:31 |
|
fun games like how do i specify my domain is it DOMAIN\\username or DOMAIN\username or DOMAIN%5cusername or none of the above
|
# ? Nov 19, 2014 22:32 |
|
or maybe its like nuget where there is no right answer you're just hosed
|
# ? Nov 19, 2014 22:33 |
|
Bloody posted:fun games like how do i specify my domain is it DOMAIN\\username or DOMAIN\username or DOMAIN%5cusername or none of the above I don't understand goofy Windows poo poo, take your computer to geek squad or run defrag or whatever What is even asking you for your domain?
|
# ? Nov 19, 2014 22:37 |
|
do development on a real operating system in a virtual machine
|
# ? Nov 19, 2014 22:38 |
|
OldAlias posted:do development on a real operating system in a virtual machine thats the comedy option where i first encountered DOMAIN%5cusername.
|
# ? Nov 19, 2014 22:40 |
|
this turned out to be the surprisingly sane DOMAIN\username:password@proxy:port
|
# ? Nov 19, 2014 22:40 |
|
QuarkJets posted:Like who the gently caress thought it was a good idea to return an int8 if you add an int32 to an int8 matlab uses the datatype of the first operand in all operators, its one of the less insane decisions they've made
|
# ? Nov 19, 2014 22:44 |
|
hobbesmaster posted:matlab uses the datatype of the first operand in all operators, its one of the less insane decisions they've made No, that's pretty stupid behavior. Other languages don't do this precisely because it's stupid. And even if you find another language that does this stupid thing, I bet it doesn't do it silently
|
# ? Nov 19, 2014 22:48 |
|
stupid and insane are not the same thing php's == operator not being an equality operator is insane for example if you want something crazy everything involving strings in matlab is a good candidate. cell arrays being separate from arrays and a 1x1 cell array containing a string being completely different from anything else or something? i haven't used matlab extensively in years and want to forget all the ridiculous stuff
|
# ? Nov 19, 2014 22:52 |
|
Fair point, the behavior is at least consistent.
|
# ? Nov 19, 2014 22:58 |
|
there are several (many?) ways to concatenate strings in matlab. strcat(), the explicit function for doing so (because "hello " + "world" just gives you an error), removes trailing whitespace. if you want to build a string containing numbers (like, in a normal language, you could do "There are " + x " foos in " + y " butts.") you have to wrap those x and ys in num2str() and if you try doing it in strcat() you wind up with "there arex foos iny butts" so you have to go use sprintf('%s%s%s... instead. alternately, stuff things into cell matrices or arrays or something and use some concatenation thing on those i dunno its retarded
|
# ? Nov 19, 2014 23:06 |
|
Bloody posted:there are several (many?) ways to concatenate strings in matlab. strcat(), the explicit function for doing so (because "hello " + "world" just gives you an error), removes trailing whitespace. if you want to build a string containing numbers (like, in a normal language, you could do "There are " + x " foos in " + y " butts.") you have to wrap those x and ys in num2str() and if you try doing it in strcat() you wind up with "there arex foos iny butts" so you have to go use sprintf('%s%s%s... instead. alternately, stuff things into cell matrices or arrays or something and use some concatenation thing on those i dunno its retarded Yeah, strings are fine on their own String arrays result in inconsistent and sometimes bizarre behavior. For instance, one MATLAB feature might return a 2D character array, whereas another might return a Cell array where every cell is a 1D character array (which is more reasonable than a 2D character array, but has to be dealt with in a different way)
|
# ? Nov 19, 2014 23:10 |
|
And I hope that you're not trying to read character data mixed with numerical data from a text file, because then you're just hosed
|
# ? Nov 19, 2014 23:12 |
|
|
# ? Apr 17, 2024 23:15 |
|
i love that blog
|
# ? Nov 19, 2014 23:13 |