Matlab

>>8517401
It's shit. People who like it have Stockholm syndrome.

>>8517401
Broadly it's better for engineers/other and Mathematica is better suited for mathematicians but it depends on the specific task, of course.

It's not free software so it's shit.

However I cannot stand having to write portable code for both Octave and Matlab so I use it still.

Also apparently they have some really good optimised code under the hood but who knows?

Very powerful, very easy to use, used so widely its pretty much an essential skill to have on your CV.

The licenses are expensive, but your uni/workplace should pay for them.

>>8517408
What about Sage? I'm currently using Mathematica but would prefer a free alternative.

>>8517401
eh, python is much better and faster

>>8517414
>It's not free software
If youre studying at a university it should be free. I got it for free at least, but it might just be participating universities. If youre at a major one you should be good.

>is it a meme
no

>too expensive
for individual people yes, but 99% of people who need to use matlab will have access to it through an organizational or institutional site license

if that's not the case for you, just fucking pirate it

>SciPy
just having the lapack libraries running under the hood isn't enough to be competitive. the nice thing about matlab is that it's got those libraries along with JIT compilation and nice paralleling support as well as lots of convenience functions and syntactic sugar to make working with matrices nice and easy. also all those toolboxes.

also the documentation is pretty nice. better than a lot of python libraries i've used

negatives: while matlab technically has support for the OO paradigm, the syntax really sucks and there aren't data structures that work well with it

if you're some fucking NEET loser who doesn't have a university or company to pay for your license then sure, get scipy instead of matlab

lol fucking loser

>>8517429
octave and scilab would both be ok for you

If you are an engineer in school MATLAB will always be used in your courses:
programming in MATLAB
numerical methods in MATLAB
stats/prob with MATLAB
controls with MATLAB
your courses (fluids, thermo) will also include significant MATLAB programming as part of the homework
many people do nothing but code in MATLAB full time as engineers

>>8517401
Matlab is industry standard whether you like it or not.

Piratebay licence 0 €

>>8517440

What I meant is that it's does not come with a free license. Free as in freedom.

>>8517401
SciLab is open-source, free rewrite of Matlab.

>>8517741

Industry standard? What industry?

MATLAB is designed to let people who have no business programming write programs

>>8518687
THIS

learn FORTRAN or C# ffs

File: 1469649398486.jpg (8KB, 236x235px) Image search: [Google]

8KB, 236x235px

>>8518687
>>8518697
Does Fortran or C# auto-parallelize your code? No? Then fuck off

>>8518713
>vector/matrix based operations
>auto-parallelized

top kek lad. leave the programming to the big boys

>>8518721
>implying other languages have things like logical indexing that map directly to SIMD operations without autistic bullshit
>implying I'm not a master's student working on artificial intelligence

>>8518733
keep writing your high performance code in Matlab, see how that works out for ya

>>8518713
The C# part was a joke. FORTRAN and C both have libraries for that, though.

Visual Basic is honestly the most useful programming language you can learn as an engineer.

most companies are stingy as fuck when it comes to buying software licenses, so most guys cobble VBA shit together in Excel because its free and everyone has it on their workstation.

>>8517401

Every time I am told to get a result in MATLAB I die on the inside and contemplate a different field.

>>8517401

I would never write a large program in python. You're just asking for trouble.

File: GNU Manifesto.png (254KB, 679x693px) Image search: [Google]

254KB, 679x693px

>>8517414
>It's not free software so it's shit.

>>>/g/oymmunism is that way

>>8518808
why's that?

>>8517748
>Free as in freedom

GPL is not "free as in freedom". BSD/MIT licenses are.

only thing good about Matlab is the command window and workspace. so convenient compared to actual programming languages that it makes going back from it painful

Is there anything that more reveals no one on /sci/ has worked in a real lab environment than when they say this "muh C/FORTRAN not MATLAB" shit

You use MATLAB for proto typing and then when you really need to get into the real shit you use C, FORTRAN or a faster language

MATLAB is used for other stuff because it's convenient. Even then, the time saved by changing to other languages is nowadays not that great since MATLAB isn't really that much slower and you can get stuff done without writing nearly as much code as you would need to in other languages

>>8519050
>I have no idea what I'm talking about so I'll spout some random shit to try to sound smart

I really expected better from sci.

>>8517401
Matlab is pretty comfortable. Easy to learn, does the job and it has sweet toolboxes. I got a license from my uni.

File: image001.jpg (124KB, 480x320px) Image search: [Google]

124KB, 480x320px

>>8517401
Freedom is inherently better.

>>8518746
But that's the thing, Matlab (and Python as well) is not meant for high-performance code. It's meant to allow people to develop tools to streamline their work without forcing them to spend half a year programming and optimizing simple taks.

It's for researchers, who have to focus on research and not so much on programming. It's all about priorities.

I mean, if you've got a biology degree and a compsci degree then sure, write all your scripts in C, but your average scientist doesn't have the time to learn C. In my experience, even Python is too intensive for many of them.

>almost 2017
>not cracking the program

why /sci/ is so pleb?? Lurk more on /g/ IT brainlets

is there ANY advantage in prototyping in matlab vs prototyping in python with the spyder IDE (or any other IDE with command and script windows and a variable workspace)

>>8519307

k but you literally have to be braindead to not understand the basics of programming

it's really not acceptable for a scientist to not at least understand python

>>8519464
"the basics of programming " to a programmer are different from "the basics of programming" to a scientist.

It would be appalling for a programmer not to have a thorough understanding of threading.

how many scientists who have made amazing contributions in biology, chemistry and physics do not have a thorough understanding of threading?

the majority.

>>8519058
I'm a grad in a lab at Berkeley

Maybe shit works differently in Europe

I get it for free every year because my uni gives it away to alumni.

Thats because its garbage that nobody wants. I just use it instead of excel because noone in my office can use it except me and they think thats amazing. Job security because i do unfamiliar things lol.

>>8519498
You're not really saving any time by using Python anyway

Hell the time savings in execution from switching to C or something similar aren't worth the amount of extra time it would take to code

>>8518808
Why not use Python to implement a C/C++ code

>>8517401
Yes, yes and yes.

>>8517408
Exactly.

>>8517414
>Also apparently they have some really good optimised code under the hood but who knows?
LAPACK and BLAS, just like numpy, R, Julia and Octave.

>>8517440
They trick you into getting used to it so they can charge thousands of dollars to your future employers.

>>8519514
No you're not, a Berkeley grad student wouldn't be stupid enough to assume I'm a Europoor.

Unless your lab is something like gender studies, you did say Berkeley right?

Also:
>Implying Berkeley is still prestigious and not a liberal shit hole

>>8520008
This is probably the worst post on /sci/ at the moment and I'm not even the guy you replied to

Either you're here from /pol/ and/or your knowledge of schools is based solely on what you read on Breitbart.com

>>8519306

Too bad communism isn't freedom and idiots can't see past catchphrases.

File: 1478673937250.gif (494KB, 387x305px) Image search: [Google]

494KB, 387x305px

>Prototyping
Use Matlab, or python if you are good with it

>Deployment
Python interface to a C++ backend

This is how shit gets done in the real world.

>>8517702
>stats/prob with MATLAB

Ever heard of R?

>>8519829
>LAPACK and BLAS, just like numpy, R, Julia and Octave.

Also Intel MKL.

>>8517401
I use it for FEA. I like it, it works but the license is expensive. I was trying to migrate to another language, any suggestions?

>>8520058
>t. knower

>>8517408
This. It's an engineer's toolbox. It's the industry standard for signal processing, for example.

>>8517401

>Redpill me on Matlab

Keep using it. You're clearly too retarded and obsessed with paying for things to use a superior language.

MATLAB is super easy to use. You can also make music and process video:

https://www.youtube.com/watch?v=dqeM0x6AWgg
https://www.youtube.com/watch?v=zzASv4G9bNA

>>8517401
Yes, yes and yes.

>>8518828
So is Matlab BSD/MIT-licensed? What is that? No? Then shut the fuck up.

You BSDkins get triggered as fuck whenever someone just mentions free software, let alone GPL.

>>8520788
Because GPL steals your freedom.

If you write code that used copyrighted code, your code is still yours.
If you write code that used BSD code, your code is still yours.
If you write code that used GPL code, your code is now owned by the "collective".

>>8520804
I'm not even defending GPL here, I don't give a shit about FOSS licenses, but Jesus H. Christ you are triggered as fuck.

>>8520813

Communism is spreading again and must be fought wherever it's found.

>>8520823
communism will invariably triumph in the end, comrade

>>8517401
>All these people talking like piracy is not a thing
Stop paying for artificial scarcity goyim

>>8520885
but that would make me a commie! and commies are degenerate! i have to pay, paying makes me a better person!

>>8520896
Stealing doesn't make you a commie.

>>8517401
MATLAB is best thought of as an extensive calculator, it lacks a lot of functionality found in other languages.

>taking a class on machine elements
>need to make a kinematic analysis program with MATLAB because that's what the engineering department uses, working on specifying cam shapes for heights at different points
>initially try to store those heights in a list and have MATLAB generate a piecewise function based on the list
>MATLAB doesn't have higher-order functions so I just have it specify them as fuckhueg vectors instead

>>8520961
>it lacks a lot of functionality found in other languages

What the fuck are you talking about?

>MATLAB doesn't have higher-order functions

http://blogs.mathworks.com/loren/2013/01/24/introduction-to-functional-programming-with-anonymous-functions-part-2/

>>8520023
>Tfw you replied to a guy who completed his undergrad at Berkeley, and left for grad school because I experienced it's downfall first hand.

My father went there too, and stories of how great Berkeley was are just that, stories and fables.

But sure keep making your clueless assumptions. Let me guess, you took a tour or two of Berkeley and thus you know how great of a school it is

This place gets stupider by the minute

>>8517401
It is okay for what it does, it lets you implement algorithms fairly easy and neither requires high levels of memory management like C or requires lots of abstraction.

It is basically C for retards (with a million algorithms already implemented) which makes it very suited for enginners.

>>8521353
>I don't know what I'm talking about

The post

>>8521449
I literally do.

while we're talking about matlab, how we we create a piecewise functions with for loops? too low iq for this

plot f(x) where x spans the domain [-7, 10] and f(x) = {2x^2, 5 <=x; 5x, 0 <= x <= 5; 3, x < 0}

>>8521466
>Uses the word 'literally' as emphasis

Claims to know what he is talking about

Top zozzle

>>8521573
function y = f(x)
if (5 <= x)
y = 2*x^2;
elif (0<=x and x<=5)
y= 5*x;
else
y = 3;
end

end

% my memory of matlab syntax is a little rusty, so put in however you properly do if/elif/else statements

>>8521607
thanks but how do i use a for loop and plot it?

>>8517401
I love using it but it's definitely not worth its price considering all the free alternatives

>>8519498
>le multithreading is hard maymay

>>8520008
>Marxist shit hole
ftfy

>>8520109
I went to a reject UC school and R was never used in the engineering department. I have never seen an engineer use R in industry.

I can chose between using python and matlab for one of my physics courses next year, what should I use? Ive taken an intro course to python and have zero experience with matlab. My uni does provide matlab for free of course.

>>8519464
It's not that it's hard to understand, it's that people don't want to spend time trying to understand it. I'm sure that any retard can learn at least basic Python techniques and principles if they apply themselves for a few hours, but nobody wants to. They just like to marvel at their collegues and be all "Oh, I could never do what you're doing, it looks so complicated!".

Seriously, I've tried to explain to my collegues so many times that the scripts I write are actually really simple, but they don't get it. I keep them so simple that all they need to do is change one or two variables at the very top of the script to do what they want to do, but they simply do not get it. Not because they're dumb (although they probably are to an extent), but because they're intimidate by the fact that 'it's code'. Simply seeing lines of code is enough to turn people off.

It's basically the same reason you can't teach your grandpa how to use a computer. It's not that he's dumb, it's that he sees unfamiliar technology and he just assumes that it'll be too hard for him to grasp (or that he tells himself this to avoid putting in the effort to learn what a folder is).

>>8521843
That's because R is for statisticians.

>>8521910
Use both. It's better to master many skills.

>>8521910
Depends very much on where you are heading in physics. In my fie (experimental particle physics)ld, python is very dominant, nobody uses MATLAB around here. However in other fields (usually more applied fields) MATLAB is very widely used.

>>8519462
can someone have a go at answering this?

>>8522073
Matlab has cool toolboxes.

>>8517401
Godlike for linear algebra
I mostly use Excel when I want to graph something though. It's nice having premade tables.

Im a senior EE and matlab makes life 10000x easier when using any kind of sampled data we've had to anaylize. I also torrented R2016 with about 2 minutes of looking online with every toolbox for looking. Anything above 2014 from my experience makes great plots also

>>8520629
lmao wtf is this senpai

>>8521573
x=-7:0.01:10;
f=[0.*(-7:.01:-.01)+3 5.*(0:.01:5) 2.*(5.01:.01:10).^2];
plot(x,f)

>>>/test/

>>8517401
It's built around doing math so the syntax is much nicer than say python that has math hacked in.

matlab: A = [1 2 3; 4 5 6];
python: from numpy import matrix; A = matrix( [ [ 1,2,3 ] , [ 4,5,6 ] ] );