[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math


View post   

File: 70 KB, 1280x720, jinrui_wa_suitai_shimashita-09-watashi-uneasy-comedy-reaction.jpg [View same] [iqdb] [saucenao] [google]
9228708 No.9228708 [Reply] [Original]

>R, julia and everything else that isn't MATLAB
why is this shit still even a thing?

>> No.9228713
File: 1.81 MB, 720x1280, metal gear solid 6 leaked footage.webm [View same] [iqdb] [saucenao] [google]
9228713

>MATLAB and everything else that isn't R, julia
why is this shit still even a thing?

>> No.9228720

>uses bloated high-level languages

Just use C/C++ brainlet.

>> No.9228721

>>9228713
pls tell me how R or julia are even remotely as good as MATLAB. I'll wait.

inb4 free
Any software is free when you want it to be.

>> No.9228723

>>9228721
pls tell me how MATLAB is even remotely as good as R or julia. I'll wait.

>> No.9228726

>>9228723
Not a piece of shit to start. Libraries/packages are easily added. None of this it's own library is not writable bs. Very good signal processing and control systems toolboxes. Actually helpful documentation and communities.

>> No.9228733

https://en.wikipedia.org/wiki/Julia_%28programming_language%29
>In 2015, the Federal Reserve Bank of New York used Julia to make models of the US economy, noting that the language made model estimation "about 10 times faster" than before (previously used MATLAB)

also doing statistics in MATLAB instead of R confirms you to be a brainlet

>> No.9228735

>>9228733
>trusting the fed
>putting up with clunky unintuitive code when there is an easier alternative makes you a brainlet
go back to watching rick and morty

>> No.9228737

>>9228735
>unintuitive
maybe if you're a brainlet

>> No.9228744

>>9228737
you <- no

>> No.9228747

>>9228708
MATLAB = overpriced shit
Slow composition of open sourced solution that costs thousands.

>> No.9228749

>>9228744
see >>9228737

>> No.9228768

>>9228708
If you ignore its horrible, horrible language, MATLAB is pretty good.

>> No.9228793

>using anything but R for stats
Write code, latex and your report in RStudio is pretty comfy

>> No.9228923

>>9228793
>Rstudio
How do you stand the response lag on the REPL? I find it infuriating.

>> No.9228939

>>9228708
>using proprietary programming languages
/sci/ confirmed for dumber than /gee/

>> No.9229045

Julia is the true patrician language

>> No.9230239
File: 173 KB, 600x355, old_hag_langs.png [View same] [iqdb] [saucenao] [google]
9230239

>be OP
>be retard engineer
>literally learn only ever touched matlab as a programming language
>make an absolutely abhorrent post on /sci/
Lmao fucking kill yourself you idiot.

>> No.9230241

>>9228708
Matlab is bretty gud, definitely better than R, but it triggers autists because it isn't MUH OPEN SOURCE

>> No.9230253

>>9228721
MATLAB is a bloated piece of shit, with narrow range of application. Just because it works for your given current needs, doesn't mean shit. It's like saying why does C++ and Java exist when there is C.

>> No.9230955

>>9228708
Matlab is one of the worst designed languages I have ever seen, it is so full of garbage and retarded Ideas.
The only "feature" it has that it has a great standard library which makes testing things and prototyping things really fast.

>> No.9230970
File: 102 KB, 534x400, 1500068778638.png [View same] [iqdb] [saucenao] [google]
9230970

> He felt for a data science meme
> I also felt for it
> feels bad man

Tho matab is shit.

>> No.9230974

Is Julia a meme? I don't want to learn another me language.

>> No.9231059

>>9228721
R is built for statisticians by statisticians and has statistical/machine learning libraries that you wouldn't even know existed

>> No.9231064

>>9231059
because documentation for it is so shit.

>> No.9231071

>>9231064
python is the only rival to R in terms of machine learning

>> No.9231085
File: 58 KB, 850x998, 1502720803804.jpg [View same] [iqdb] [saucenao] [google]
9231085

>> No.9231088

I only want to learn meme languages. I just got done learning ArnoldC. What's next? Python?

>> No.9231089

>>9231088
everything that isn't assembly, c and c++ is a meme language.

>> No.9231103

1. MATLABs language is horribly inconsistent.
2. The libraries, astonishingly enough considering they almost exclusively come from one single entity, worsen this problem even further.
3. The often praised plotting capabilities are only easy to use or even viable for the simplest of applications. They fall short in absolutely amazing ways.
4. The GUI tends to crash on complex models. I'm surrounded by people working with complex Simulink models at work and day in, day out, they have it crash on them, several times per day.
5. The MATLAB file formats are a horrible mess and there is no easy way to use them from other languages.
6. It is, of course, pay-for. And yeah, if you're an organization, that matters.
7. It is slow as shit for the heavy lifting. And yeah, I'd like to see a source for >>9231085
because I have a hard time believing that - there's not a lot of code out there than can beat the (C-implemented) SciPy and NumPy libraries in terms of speed.

On the plus side, Simulink is an amazing tool if you want engineers writing their own software and you don't care about the quality of it. Also, it does a decent job at interfacing with many things, better and simpler than most other languages at least.

Anyway, if programming languages were a competition, MATLAB wouldn't even be near the front. On top, you have still C/C++ on top of the compiled software, bare metal market; Java (and JVM driven langs) on top of the general purpose market; and Python on top of the purely interpreted languages with a lead so huge it has lapped MATLAB a few times alread. So yeah, if it were a competition, in the market MATLAB is targeting Python is already basically a monopoly, even though engineers don't realize this at all (amazingly enough).

>> No.9231130

>>9231103
>It is slow as shit for the heavy lifting
[citation needed]

>> No.9231134

>>9228708
the same reason that all kinds of shitty technology exists

because it has libraries that no one can be bothered to write into better languages.

>> No.9231154

>>9231103
matlab and the various python packages all use hte same BLAS/LAPACK libraries under the hood so the difference is probably down to how much overhead the benchmarked code had. I could totally see a benchmark like that being (possibly even unintentionally) biased to matlab

>> No.9231180

>>9228726
I agree, R is pretty good.

>> No.9231187
File: 373 KB, 1844x1536, r-matlab-julia.png [View same] [iqdb] [saucenao] [google]
9231187

>>9228721
well lets have an objective comparison

>> No.9231190

>>9231187
Rcucks on suicide watch

>> No.9231205

op here,

after actually looking at julia I take it back julia is amazing. R is still a pile of dogshit tho.

>> No.9231218

>>9231187
>IDE
>RStudio
>Not RGui
plebs please go

>> No.9231229

What does /sci/ think of Maple? They're teaching that shit instead of matlab in my uni.

>> No.9231230

>>9231205
You should take a look at R as well

>> No.9231235

>>9231229
it's symbolic instead of numerical which is matlab, tho matlab does have symbolic computing too but it's not the core of it.

>> No.9231241

>>9231235
wait, maple is both.

>> No.9231251

>>9231235
Which one is better though

>> No.9231259

>>9231251
they have different purposes.

>> No.9231261

R and Matlab are both shit, python is awesome as long as the heavy lifting is done in c/c++

>> No.9231264

>>9231251
maple is probably better for symbolic computation. for a time matlab used their code for their symbolic toolbox. tho there is also mathematica for that too. matlab is better for the numeric stuff. matlab's strength is linear algebra, so if you're dealing with matrices it's the way to go also the ide is great.

>> No.9231271

>>9230241
You do realize there is GNU Octave? A literal MATLAB(tm) clone?

>> No.9231274

>>9231271
it's not a complete clone and it's missing the parts of Matlab that are most worth using

>> No.9231291

>>9231259
...thanks for the non-answer...

>> No.9231296

>>9231291
it's not a non-answer at all. they're not competing products, so comparing them is pointless.

>> No.9231386

I use Perl for PCR funsies.

>> No.9231404

>>9228720
> spends an entire day reinventing the wheel so his code runs 3 seconds faster
> even if he's doing it because of his inferiority complex it's still not haskell or lisp
> suggesting C++ even ironically
> calls other people brainlets

>> No.9231425

>>9231404
>haskell or lisp
Worthless langs. Stop with the memes kid.

>> No.9231433

>>9231425
this was exactly my point, work on your reading comprehension.

C and C++ certainly have uses but they are not the same as Matlab's.

There is no point rewriting your code in C unless you're going to run it literally hundreds of thousands of times.

>> No.9231456

>>9231404
>3 seconds
>30 seconds
>300 seconds
>3000 seconds
>30000 seconds
And the next thing you know your shitty space invaders indie clone cant run without gayforce high-g-triple-xXx 29870 xtreme overvolt edition.

>> No.9231908

>>9228708
Wait till babby learns C/C++ and stops fearing pointers
Then you'll wonder why MATLAB/Java/etc still exist

>> No.9231916

>>9231404
>spends an entire day reinventing the wheel so his code runs 3 seconds faster
You don't know C/C++ has libraries, or are you just pretending?
>it's still not haskell or lisp
Still not a dinosaur language without native pointers in the case of haskell?
>lisp
kek@suggesting a dynamically typed language over C/C++

I know you're trolling at this point, but pls stop

>> No.9231956

>>9231296
not him but there's a difference between blunt and brief

blunt is "neither" or "they have different purposes"
brief is explaining what those differences are