Why is MATLAB so shit to use?
Other urls found in this thread:
gnu.org
en.wikipedia.org
twitter.com
works on my machine
git gud or use GNU Plot
What’s so bad about it? It isn’t supposed to be able to solve every problem in the universe well, you know
The programming language is shit. It feels natural only for those who have never programmed in a traditional computer programming language.
Software designed for shitty "engineers" and "mathematicians". Any good scientist will use a real programming language if he knows how to code.
H A R M F U L
A
R
M
F
U
L
It's fine.
Because of the moldovans
It's not meant to be used by brainlets.
It's easier than python, what are you having trouble with
what
aerospace engineer here. What are you smoking user? Matlab is good bit of kit. Can solve equations of motion (dutch roll, sprial mode all that shieeeet) and transformation functions too! I like simulink as well, you can create cool mathematical models with that.
Drop-in compatible with many Matlab scripts.
Simulink is the best tool for control engineering.
MATLAB is awful, I've been using it for 3 years and still hate nearly every convention (and lack thereof).
That guy is clearly a LARPer and shouldn't be taken seriously
>good bit of kit
this is the most boomer thing i've every heard
I used MatLab as part of a large data analysis system and it was very usable. The documentation was accessible and useful. The platform was stable. The performance was decent. Any capability I wanted, I found. So what's the problem, skro? ;)
But it is great?
What are you having issues, with?
Probably the greatest standard library out of any language, together with awesome documentation, the language itself is also quite good and you can easily translated algorithms into it.
>"boomer"
You're famous!
Is OpenGL and C++. Then you only have yourself to blame if it's shit
when i was at uni, it was the cunts who struggled to make a cone on CatiaV5 that complained about how shite matlab is.
>The performance was decent.
It is absolutely awful at certain things though, especially if you do things which aren't matrix stuff, where it can call BLAS, LAPACK or such.
Matlab is ONLY good for matrix math, Maxima is a better math programming language and is free
That wasn't my experience; my application involved signal analysis and MatLab seemed to process it all in a very reasonable fashion from what I could tell - it enabled me to keep the majority of my attention on the bigger system rather than attempting to optimize how the calculations where performed, and I think that's the point of using the platform.
>tfw did bioengineering
>used matlab for some bioengineering shit and models that was super intense made lots of 3D plots n shit
>get into mackanical engineering because don’t want to go to med school
>easier than bioengineering shit but equally as complex just a bit simpler to code n shit
>get into all kinds of control theory with it
>graduated and first good paying job was a sales engineer job
Feel like I downgraded but life is better figuring out how to use excel and word instead of spending hours on complex matlab shit that made me so frustrated
>sales engineer
do you mean salesman
Learn to read you thick fuck.
>first good paying job
Its your foot in the door. Sales engineers still have to be competent with engineering fundamentals to secure contracts and shit. I know a few who've gone into tendering and project delivery after they've gained experience.
i'm not biomed/mechanical eng user btw.
> Dutch roll, spiral mode
What about pile driver, reverse cowboy and DVDA?
Some points in no specific order:
- Index starts at 1
- Anything that requires slight deviation from vectorized functions requires either slapping together multiple clunky statements (obvious decrease in performance), or you can write it yourself and get nowhere close to the vectorized performance
- You have to pay for parallel processing
- Horrible names for functions ("eye()" for identity matrix)
These are just a few that came to mind, but if I write some code, I'll be sure to find some more annoying aspects.
t. Matlab employee
>mackanical engineering
>thick fuck
What does packing bong cones with weed got to do with Matlab?
>shit to use
brainlet
LARPer
Actually, I am an agent of the Illuminati involved in a post Cicada 3301 recruitment campaign looking for intelligent, rational technologists to work towards a Brave New Wold Order.
*World
>starts indexing at 1 not 0
So hard to get used to
You are not the center of the universe. If you find something that seems foreign to you, that doesn't mean it is universally wrong or bad or not useful. Why haven't you figured that out yet?
>Why haven't you figured that out yet?
Isn't it obvious?
So you're saying it is actually used by intellectuals? Not some pajeet code monkey? Barry deep
Everything I said was objectively true; nobody likes indexing from 1, it's not even foreign, it's just against convention. Don't get your panties in a bunch just because I insulted your favourite language.
Indexing from 1 is from its matrix/ math background (matrix labatory). Again, its user base are intellectuals not code monkeys. Adjusting algorithms to work for such indexing is nothing and can be automated.
>LARPer
want proof?
java
> its user base are intellectuals not code monkeys
>Implying I don't use it daily
>Indexing from 1 is from its matrix/ math background
Still doesn't change that its a programming language that uses 1 as the starting index. Additionally, lots of math use 0 as the first index, so your point isn't fully convincing. This isn't even my worst gripe about the language, it's just the most relatable one; I still like using it.
It sucks I don’t use any of the academic skills really or do the same kind of R&D work that I had sought out to originally do but that mindset and experience gives me a leg up as a sales engineer (technical sales). We got lots of folks that deal with pure marketing/business shit and I’m sort of a mix.
>tfw future CEO
>nobody likes indexing from 1
*You* don't like indexing from 1. You are not everybody.
>signal analysis
Techno-spook!
LOL! Bang on analysis, my man.
How does this board attract all of these whining little retards?
>This is the only point you can make
*More people than not*
How did you come to that conclusion? Does it just "feel" right to you?
This.
No, I think the numbers speak for themselves.
Fellow aero here, when I don't feel like using Nastran I use it to calculate interlaminar stresses on a composite plate under different loading and constrainment cases.
>this much denial
Git gud
Python is overall better but matlab has tons of complicated things it can do in a snap. And I mean a lot.
That's goofy. You're just a butthurt retard.
>"eye()" for identity matrix
Holy shit I didn't know this.
If you mean to say SciPy is nifty, then yeah, it is but I wouldn't trust the results for anything serious.
>Showing proof that people prefer X over Y
>Im butthurt because the stats support my claim
LMAO
MathWorks revenue~$900 million for 2017. Plenty of people not only us it but are willing to pay to use it. So many people use it there is even a thriving FOSS clone - GNU Octave.
You're just spinning in an attempt to justify your preferences and your desire to project your preferences onto everyone else. This is egocentrism which is a developmental disorder if you over about six years old and living in a modern society.
>I need to justify my preference
>Getting visually rustled over the fact that people like one thing over another
I'm not justifying my preference, I'm proving my claim that people prefer 0 indexing. You're sitting here like an armchair psychologist trying to analyze my life because I said something you don't like about a language. Quit coming up with excuses.
We're trying to help you, skro.
MATLAB has a FEA toolkit?
:P
>code monkey retard talking shit about engineers
fuck off, you ape
>>matlab is shit
t.brainlets
Matlab might be not so special like they used to. There is some functions that other lang like Python manage to come close to already.
However in linear algebra, statistic and sciences data management it is beautiful quite clean and easy to use.
Matlab plot is beautiful and easy to customize out of the box.
Simulink control system design is number one of its kind
I like Python and C more but cant deny that Matlab has it own place in the industry
>>get into mackanical engineering because don’t want to go to med school
>mackanical
>expecting engineers to be literate
nice damage control bubba, go back to watching wwf
if you aren't literate as an engineer, you aren't an engineer
why? They're not lawyers, engineers get things done and they dont need perfect spelling to do that
not that guy, but I work in finite element analysis and optimization and it most definitely is capable of FE methods
there's the PDE package, which is capable of using FE methods to solve 2d and 3d problems, but our team only uses handwritten in-house code, so i don't know how reliable it is.
I dont think I've heard of anybody doing any serious FEA work with matlab. Maybe thats changing
>our team only uses handwritten in-house code
Whats the story behind that? I'm guessing you work on similar problems or the boss doesn't like paying for licenses
>I don't think I've heard of anybody doing any serious FEA work with MATLAB
sorry, i should have clarified i'm working at a university - it is by no means an industry standard, although it's capable of rapid prototyping and being pretty fucking fast, thanks to BLAS and automatic multithreading. but there probably are not many serious companies who use it for FEA, no
>I'm guessing you work on similar problems or the boss doesn't like paying for licenses
well, not only our university can't afford all of the MATLAB packages, we're using it in multiphysics problems and some novel applications, so we're constantly changing the code to, say, implement fluid-structure interaction or some other optimization algorithm.
it also helps that we're (so far) only using it for 2D examples, so the code is pretty small (about 300 lines at most) and easy to maintain, so it's easy for an undergrad to understand
It makes sense from a math perspective though. "I(n)" as a symbol and all that.
Hes a jelly ruski pig
Just don't treat it like a OOP language. It's functional programming with some fancy matrix multiplication.
>- Index starts at 1
Should be common sense for all "higher" level programing languages, this makes it significantly easier to write mathematical algorithms.
>- Anything that requires slight deviation from vectorized functions requires either slapping together multiple clunky statements (obvious decrease in performance), or you can write it yourself and get nowhere close to the vectorized performance
Yes, it isn't a general purpose language, who would have thunk that MATRIX LABORATORY was a language primarily focused on Matrix and vector computation?
>- You have to pay for parallel processing
Just like for the rest of MATLAB
>- Horrible names for functions ("eye()" for identity matrix)
Nothing wrong with that.
Indexing from 1 is common sense, at least if your default type is a MATRIX which always are indexed starting at 1.
because it was designed for what it does. It trades approachability for power and depth... because you only have to approach it once.
People like standard interfaces, because they skip the acclimation phase. A calculator has a standard interface. Matlab is not a calculator, expect to have to invest in return for power and flexibility.
I recognize that bulge
You speak of mathematical models transformed to lambda calculus rendered in one of a few rendering libraries (probably OpenGL due to its age).
>Dilbert
t. salty boomer
Better make that transition sooner rather than later in your hypothetical trajectory you see yourself going. Sales is a trap where you soon realize you need more social than skillful knowledge to deal with international clients. Usually you're given a sector of the world or a country (generally dictated by language). You need to know how those people might act, their mannerisms,and whatnot.
The sooner you transition, the happier you'll be if you truly want to be involved in the *real* operations.
Only core Matlab and also far slower. More like a free substitute than a replacement. Nevertheless is fine for simple use cases.
>that guy that thinks Maxima is suitable for numerical computing because it has some related functions
>Everything I said was objectively true
Stop being a spoiled child.
>nobody likes indexing from 1
I don't like 0 either. I simply don't care.
>it's just against convention
What that convention will be? Most if not all math/science related programming languages use index-1: Fortran, Matlab, Julia, IDL, APL, Wolfram, Maple, R. Even some general purpose languages use index-1: Smalltalk, Algol, Erlang, Lua.
Yes.
It's not.
>It's functional programming
Just because it has anonymous functions it's not functional programming.
It's one of the best designed UIs around though...?
You freetards use Scilab?
fuck off
Just got home, stand by ya fucking chump
Very nice! I used ansys for thin plate bending when i was studying, I would have used matlab if I was allowed
scribbled on the emblem/crest to negate google image searches
>Spends actual minutes staring at keyboard trying to figure out how to index
>Doesn't know shit about signal processing
MO-OM! MATLAB SUCKS!