/sci/ - Science & Math

File: 1.94 MB, 640x640, 1599011426877.webm [View same] [iqdb] [saucenao] [google]

Anonymous Sun Sep 13 21:35:08 2020 No.12117143 [Reply] [Original]

MATLAB/Octave vs Mathematica vs R vs Python vs Haskell etc.
Does anyone use any of these for anything?
Is there any reason to learn proprietary scripting languages or it is better stick with Python/Haskell? For example you need to solve a quadratic or cubic equation. Does any of these math packages do it out of the box or would you need to write a script? Or it is easier to use online calculators? Which of these packages is best for visualizing linear algebra/matrix transformations etc? Is there something built in or you'd still need to write a bunch of scripts?

>>	Anonymous Sun Sep 13 23:41:14 2020 No.12117527 Matlab's algorithms are more efficient than Python's in certain instances. I've seen this first hand during one of my projects. I haven't worked with the others.

Anonymous Sun Sep 13 23:47:23 2020 No.12117549

>>12117143
Octave is not propriety. Python can be cumbersome to deal with compared to octave or matlab to do the same thing . I use matlab/octave mostly for visualizing data and to do quick math and it’s definitely less cumbersome then python. Probably the same with Haskell although I’m not familiar with it. To graph 2 matrixes of data it’s literally plot(x,y) and a graph pops up no imports or anything. While python of course is of course superior in many different ways it doesn’t beat the ease of use with matlab/octave. It can do all the symbolic shit too easier. It’s also useful if you ever end up doing control systems to know your way around matlab since it’s integrated into simulink.
Basically you can do everything in python with all those libraries you can do with matlab/octave but easier. One of the most notable things is better error messages.

>>	Anonymous Sun Sep 13 23:48:25 2020 No.12117555 >>12117549 I mean matlab and octave are easier

>>	Anonymous Mon Sep 14 07:43:41 2020 No.12118665 >>12117555 Yeah, thats been my first impression. Python is more powerful but it is almost as if you tried to build your own "matlab".

Anonymous Mon Sep 14 08:01:35 2020 No.12118715

Another thing I just realized I don't really need a fully fledged CAS. All I probably need is basically a fancy graphing calculator. That reduces the list down to matlab vs octave I guess. But if anyone cares about a CAS, there is also Sage which is python based.
That said, there are plenty of online options for numeric stuff, so I am not even sure I need to install anything on my computer lol.

>>	Anonymous Mon Sep 14 08:12:51 2020 No.12118746 >>12117143 Just use Coq, it's cumbersome but at least you are sure that your program is correct.

Anonymous Mon Sep 14 08:18:57 2020 No.12118759

>>12117143

>MATLAB
is proprietary garbage that you should only use if someone is paying your license and want you to specifically write MATLAB programs.
>Octave
is a meme.
>Python
is the most flexible out of those, and will get you the farthest if you're not wanting to learn more than a single language.
>R
is really good at statistical analysis, but not at anything else.
>Haskell
If this was useful to you, you wouldn't be asking for advice on what language to use.

Just go with python. Then learn a compiled language like C/C++ or FORTRAN down the line if you need SPEED.

An alternate recommendation would be Julia, which is an excellent language. However, it is still in its infancy, and is still somewhat of a hipster language. (but will likely become mainstream in a few years given its backing.)

>>	Anonymous Mon Sep 14 08:20:27 2020 No.12118763 >>12117143 python >> haskell > matlab = mathematica >> r

>>	Anonymous Mon Sep 14 08:48:42 2020 No.12118837 >>12118759 >R for statistics Looks like it has a bunch of other stuff and technically should be able to do everything that matlab/octave can do, but I am not sure

>>	Anonymous Mon Sep 14 08:49:44 2020 No.12118841 What about Maple or Magnum?

Anonymous Mon Sep 14 08:51:01 2020 No.12118845

>>12118759
also
>learn a compiled language like C/C++ or FORTRAN down the line if you need SPEED.
To use with what package? Do you mean use a standard C/C++ compiler plus 3rd party libraries or the other way around, some math package that allows you to write extensions/libraries in c/c++?

Anonymous Mon Sep 14 09:01:37 2020 No.12118877

>>12118837
Yeah sure, but its not particularly useful for general purpose stuff (neither is MATLAB/Octave). If you're autistic enough, you can do your computations in LaTeX or powerpoint as well (both are Turing Complete), its just not a very good idea.
But seriously, out of those languages Python will offer you the most value for your time on account of its widespread use and flexibility. Autists will convince you its terrible (and they'd be right, for technical reasons), but its still the best all around language today. Numpy, scipy and matplotlib covers most of your numerical needs.

>>12118845
sometimes, the problem you're trying to solve can't be "vectorized" and solved using numpy (which is a wrapper for fast C code). So, you're left having to write your own loops in python (same applies to matlab, R and octave).
At this point, it MAY be a good idea for you to speed up these performance critical parts of your program by writing your loops in a fast language. But this will only be an issue if you're interested playing with larger/complex problems. So basically, don't worry about this for now. You can still use python for these things, it'll just take longer to get your results.

>>	Anonymous Mon Sep 14 09:02:15 2020 No.12118881 >>12118845 With Python you dumbass. You write the packages for it in C or use ones that already exist.

>>	Anonymous Mon Sep 14 09:03:51 2020 No.12118885 >>12118877 >for general purpose stuff like what?

>>	Anonymous Mon Sep 14 09:11:34 2020 No.12118909 >>12118885 Like literally anything you'd want to program. Python has a fucktonne of packages and you can do basically whatever the fuck you want in it in addition to solving maths problems. It's basically the swiss army knife of programming languages.

>>	Anonymous Mon Sep 14 09:48:26 2020 No.12118990 >>12118837 >>12118759 R is awesome if you are working with linear algebra. If not, there are options to write compiled C++ functions which could be called right from R

Anonymous Mon Sep 14 10:23:34 2020 No.12119063

>>12118990
Matrix/vector syntax seems to be more cumbersome in R compared to matlab/octave.
You have to use c('...') or something like that, and then some extra parameters when defining a matrix. And operators look like %o%. matlab seems more intuitive, it is just [...;...] and a more natural operator syntax.

>>	Anonymous Mon Sep 14 10:33:47 2020 No.12119087 File: 327 KB, 859x960, KoAvFZk.png [View same] [iqdb] [saucenao] [google] >>12117143 >Yes, I use MATLAB, why do you ask? >What do you mean by "proprietary garbage"? Sorry I don't understand poor speak

Advanced search
Text to find
Subject [?]Search by post subject. Leave empty for any.
Username [?]Search for user name. Leave empty for any user name.
Tripcode [?]Search for tripcode. Leave empty for any.
Email [?]Search by email. Leave empty for any.
Filename [?]Search by image filename. Leave empty for any.
From Date [?]Enter what date to start searching from. Format is YYYY-MM-DD
To Date [?]Enter what date to start searching until. Format is YYYY-MM-DD
Image hash
Search in	All Posts OPs Only
Deleted posts	Show all posts Show only deleted posts Only show non-deleted posts
Internal posts	Show all posts Show only internal posts Show only archived posts
Order	New posts first Old posts first
Capcode	All Posts Only by Users Only by Mods Only by Admins Only by Developers
Results	Posts Threads
Action	[ Simple ]