Thread 106488813

why the fuck do they still teach this shit when python and R exist?

>>106488821
Genuinely baffling. Industry inertia?

>>106488821
Good docs, simulink, and not having to deal with freetards when you have a problem.

Gnu Octave won.

>>106488813 (OP)
>it just works
>used by any serious company for tech
>no freetards
>no indians (they cant afford it)
>readable
>no retarded "but my way is better" autists
>plenty of toolboxes that just work
>easy to use
guess it's that simple

What were you trying to do?

How did Mathworks drop the ball on neural networks so badly? They still don't have tools comparable to Keras, which is what I'd expect Matlab NNs to be. They're trying, yet so much is missing. But more than that Matlab's a linear algebra beast with dead simple GPU integration, so how did they fall so far behind python?

>>106488813 (OP)
Of course I wouldn't be able to use it today, but the computer arithmetic course at university was the only time I was good at maths (and I it was fun).

>>106488821
Python and R can't compete with the documentation and the sheer number and completeness of the toolboxes. Also if you can manage to follow general best practice it is fairly performant because most of your processing time will be spent in their optimised toolbox functions.

That was fast

>>106488821
Better visualization.

>>106488813 (OP)
why would i use this instead of mathematica?

>>106490159
Based. I’ll defend Matlab against the freetards, but Mathematica is the goat. The argument for Matlab is probably certain domains where you’re using simulink, but I don’t have experience with Wolfram’s similar product SystemModeler to say whether it’s better.

>>106488813 (OP)
>a 40 line script cannot take a lot to run
are you retarded?

>>106488821
The non-meme answer is inertia. It's worse than Python and R in every way.

>>106488813 (OP)
I can make 10 line script that will take days to finish, what your fucking point?
Also, unless you're using specialized modules like Simulink what's the use-case for this proprietary trash compared to Numpy?

>>106489284
>no indians (they cant afford it)
>neet didn't know about uni provided licenses
Lol

>>106490351
>proprietary trash
I instantly lose respect for Stallmanites when they talk this way.

>>106488813 (OP)
I wish they had forced us to learn Mathematica instead of Matlab in uni. Matlab is just a worse python in my experience while as Mathematica seems actually useful and distinct from toy scripting languages given its computer algebra capabilities.

>>106490159
>>106490333
>>106490661
Stephen Wolfram is an incredibly autistic retard that wants everyone to use Wolfram Language for everything. Wouldn't be surprised if the Mathematica installation is like 20GB by now with all the bloat, and outside of some technical computing where it's genuinely useful, it's a hobby language at best.

Why the fuck does Matlab not have static explicit types (or whatever they're called). I think like 90% of my time being frustrated with this software was just trying to wrestle with whatever all kinds of stuff I can stick into the functions and what they can return. God Matlab would be so much fucking better with types.

>>106490941
everything would be better with types for the small price of a bunch of zoomers foaming at the mouth

>>106488826
>industry inertia
more like "industry shilling"
Industry gives free licenses to students and universities just so that people get used to the tools, which will automatically get them into workforce in their specialized workflow

>>106489284
>>it just works
Limited for anything more niche than a tutorial from their page use
>>used by any serious company for tech
These so called "serious companies" are all bloated old companies where people don't do shit for weeks and don't get fired
>>no freetards
you obviously don't know what you're talking about since Octave exists
>>no indians (they cant afford it)
See above, + cracks exist. You're dumb
>>readable
if you did anything in life you would know that python is as readable as matlab
>>no retarded "but my way is better" autists
you never went to matlab community forum huh?
>>plenty of toolboxes that just work
that box you in, and if you need anything more or specific then you're fucked
>>easy to use
just like python

You either just started your uni work or you never did anything real

>>106488821
The huge library of function, gnuplot looks better and it's excellent but matlab has everything plus documentation in a single package.
Octave by comparison is a clusterfuck and comparatively without documentation.

>inb4 matlab or octave are slow
Yes, but most problems are related to people using C-like programing instead of more vectors

>>106488813 (OP)
saar*

ITT : /g/ criticizes yet another technology worth learning to get a high paying job

>>106491373
Can you actually land a good job with MATLAB?

>>106490941
They're getting better at types while staying dynamic
https://www.mathworks.com/help/matlab/ref/arguments.html
it's definitely the worst part about looking at some other brainlet's code. Using size constraints on arguments is a godsend too. Keeps me sane and avoids the temptation to overengineer.

>>106491512
Only if you're an engineer not a software engineer. If you're worried about AI taking your job don't bother with Matlab.

>>106491184
>t. i only used matlab for vectors
matlab has specific real life integration out of the box that python can only dream about, I can program a real system, hook it on matlab and it does it out of the box, on any other language you need a shitton of middleware slop
>python as readable
python is a programming language that needs an IDE, matlab provides an IDE
>no indians
this x1000, indians flock the open source free slop, matlab is out of their league
>toolboxes
they are extensive enough for any need

I know you're brown, give it up and claim Jupyter is just as good, no need to overreact

>>106492516
>I know you're brown, give it up and claim Jupyter is just as good, no need to overreact
Does Matlab even have the equivalent of notebooks, or are those somehow "useless"

>>106492957
> Notebooks
https://www.mathworks.com/help/matlab/live-scripts-and-functions.html
> Jupyter
Yes, they reworked the whole ui so it runs in browsers. I hate it because figures are migrating to the newshit and old ui elements are going away. But you get 20 hours a month of free matlab in the cloud, which is actually useful and works really well in browsers.
https://www.mathworks.com/products/matlab-online.html

>>106492957
yes, mlx or live script, allows you to run portions of the script, insert images and more
you can also run scripts easily in matlab, convert .m code into c++ and have it more optimized than pycuk will ever be

>>106488813 (OP)
there literally is no alternative to simulink

>>106492516
Only retards need an IDE.

>>106494560
t. nocoder

>>106488821
python and R are still inferior when it comes to usability by techlets. Every other midwit can use Matlab, but not python or R.

>>106488837
>not having to deal with freetards when you have a problem
Yeah if you have a problem it will never get solved. The entire idea that paid software outsources your problems to the company you pay is a farce.

>>106494856
Verifiably untrue.
>>106494842
Holy cope. You are not smarter because you took a more complicated path.
>>106494560
Retard
>>106490870
Wolfram is a Genius, you’re some retard on /g/ whining about muh bloat because 20 GB is a lot of space on your shitbox from 2003. Your estimate is curiously close though, 7.3 GB for the base install, 11.3 GB for the documentation. Add in some of the on-the-fly downloaded datasets and you’re at 20.

>>106488813 (OP)
>block you're path

>>106495807
I went to a leaf university and they were pushing this hard, sorry not sorry, but I used mathematica instead. I know, I know… it’s just that I used mathematica is all. HAAHHAAHAHAHAH

The real reason is that I already knew it. I had pirated it years earlier when I was 15, solely because I thought the logo looked cool, without knowing what I was getting into (also the case of the modern NatSoc — many such cases); now I’m addicted to it. Math, not even once.

>>106493646
>onvert .m code into c++ and have it more optimized than pycuk will ever be
https://numba.pydata.org/

>>106488821
Python is fucking gay

>matlab
Install a huge package and everything just werks. Manly choice
>python
Everything is fragmented and you need to take extra care of virtual environments to not break stuff. Utter gay

>>106495873
I didn't wanna pay for that shit so I googled alternatives and started using xcas. I think I did pirate it too but that always sucks.

>>106497190
>I didn't wanna pay for that shit so I googled alternatives
same.
And I settled on gnu octave but lately.
I once felt the urge to create a matlab where the script is TeX. Create a custom TeX Engine and it didn't end well.

>>106497217
Octave for matlab, xcas for maple.

>>106497389
I'm not endorsing it, but the poor man's Mathematica is SageMath.

>>106489700
>Boomer: MATLAB Perfect
That's it

>>106494560
Sure, if you just wanna rice your desktop and write trivial C programs.

>>106497867
>Software just works
>Zoomer: Ughhh it's boomer!

>>106497867
Sadly this. Engineer here. I remember us being forcefed MATLAB as the first go to for all the linalg and optimisation problems from subjects. But it's an abortion if you cross a certain level of complexity of problem in terms of organised structure. There have been better options since then. And yes, it's boomer tier. Gen X here.