Is there a language more perfect than Python for Veeky Forums coding?

Wish granted, you now have to use PHP for the rest of your life.

Ok, not that one

That's fine, you're now a Ruby on Rails developer.

FORTRAN

Grandpa, what did we tell you about trying to use computers, you know they only make you confused and angry. Come back to the living room, we put your Matlock season 3 VHS on.

PHP isn't that bad.
t. somebody who has actually used PHP in a professional development environment
JS is way more nightmareish

There's literally nothing wrong with javascript.

*blocks ur path*

>matlab
matlab is becoming obsolete.

matlab is proprietary.
Python & R (for statistics) are better in many ways

Mathematica is quite good, but also expensive and proprietary.

>(((lisp))) , (((scheme))) , (((racket)))

(((lisp))) created by (((John McCarthy))), Math PhD
Son of a Irish father and a (((Lithuanian Jewish mother))).

>((( en.wikipedia.org/wiki/John_McCarthy_(computer_scientist) )))

Is Python the best language to learn as a MechE?

If you stop using everything created by jews you're going to end up not having very much left. He's not even full jew, you're taking the /pol/ thing too far.

(lisp (created by
(John Mccarthy
math PhD son of a (Lithuanian Jewish Mother)))

It's funny though, right? Because lisp uses lots of parenthesis. Get it?

>Is an interpreted language a not so bad language to learn as a person in STEM?
No.

>implying R is comparable to Python

It's so fucking comfy dude

It is because that's what Python is actually used for most of the time, data science shit.

>C++ is /v/

>web apps
is that way

If by python you mean numpy, scipy, and tensorflow then no, there isn't.
If those were out of the picture then I'd probably use Matlab, just for the convenient toolboxes

>python is king of ... web apps ...

Jupyter Notebook is very comfy
jupyter.org/
>he Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text.
>Uses include: data cleaning and transformation, numerical simulation, statistical modeling, data visualization, machine learning,...

enjoy having your code break on you when you copy and paste it from one editor into another.
python is pushed forward on the power of memes, nothing more.

slow is fine, Veeky Forums fags are not writing systems, you are using high speed functions written in C++ using all the fancy parallel libraries to do your little computations

>C++ is /v/
I don't think /v/ knows how to code, they claim to built game engines, but when asked to backup their knowledge they are unable to explain the most basic things about programming.

Idris, duh

True Veeky Forumsentists use the untyped lambda calculus for Veeky Forumsentific computing

>not using a language where you can write code then reason formally about it (as well as automate said reasoning) in the same language
What's your excuse Veeky Forums?

unironically no there is not

> does element wise operations on python lists using for loops
>hasnt heard of numpy

>"gee whizz isnt python super slow?"

>uses a library coded in C and just calls it from python
>s-see, python isnt slow!

>>uses a library coded in C and just calls it from python
>>s-see, python isnt slow!

the main python interpreter, cpython, is coded in C. we may as well say python is just a C program cause it is. every statement in python executes stuff coded in C.

but then the language python isnt C, its a high level language that has powerful features that are fast because their underlying functionality is implemented in C. this is a good thing. this is a feature. python provides ways to combine C code into powrful and fast python libraries. but when i use numpy i use python, not C. btw there is also plenty of fortran wrapped up in libraries. C++ does similar things, so do many other languages.

think about that a lot before you reply with a snarky reply

>using a theorem prover developed by microshit
i shiggy diggy

are you mentally retarded? when people say C++ is fast, they don't mean using the libraries written in fortan is fast, they mean running code written in that language is fast, you (or that guy) are saying that python isnt slow, because there are libraries that are fast written in another language, thats not how it works. Python gets compiled into C? great for it, everything gets compiled down to machine code so I guess there's actually no performance difference in any language ever made.

>when im using numpy im using python
depends completely on what you are doing, a long complicated program using numpy you can argue is python, a loop over an array (like that post was talking about) is entuirely done in the underlying C, not in python. By thil logic you could write an entire program in C, wrap it up in a method in Haskell that just 3 lines of code and claim your ACHTUALY using Haskell.

>b-but its a high level language!
the chain of posts were about if its slow or not, now you bring up the fact that its easier to code in it than C? Why?

numpy is a high performance numerical library for python. apparently it doesnt count as python because its partially implemented in C. well cpython is implemented in C, so i guess python doesnt count as python because its implemented using C.

or we can count numpy as being python, especially because its API is pure python, just like python is pure python.

the only people worse than pythonistas are rustfags

protip guidos: nearly every language in use that isn't C can call C shared libraries because the C ABI is basically universal

>1) nobody uses lisp
I use elisp...

step aside, best language coming by

kek

language != implementation

ARE YOU SEVERELY MENTALLY IMPAIRED???
When people say C++ is fast they mean that compiled BLAS/LAPACK libraries under an INTEL COMPILER and after MUCH LOW LEVEL ASSEMBLY TWEAKING, are fast.

Jesus fuck, you even acknowledged that C gets compiled to machine code, so in the end WHAT IS THE FUCKING DIFFERENCE? Just the compiler and how good it is at making automatic optimizations.
There ARE NOT fast languages, just better compilers.

You are as insufferable as these CSshit codemonkeys who know nothing about numeric algorithms at all and make "le kewl lorentz attractor" with a forward euler integrator. If you don't know math, and you don't know computation, stay clear of computers.

>not coding in assembly

>There ARE NOT fast languages, just better compilers.
No. What optimizations are available to a compiler in a large part depends on the semantics of the language. What the compiler can prove about your code is different from what you know about your code. This is why static typing is popular and why many languages offer metaprogramming.

>There ARE NOT fast languages, just better compilers.
>Taking my strawman and running with it
kek

For scientific programming, it's not only comparable to Python, it's superior.

Scientists don't need a general purpose computing language. They're never going to have to design huge server applications or web platforms or anything like that. They need to analyze their data. What scientists need for that is a language specifically designed around making the manipulation and analysis of data trivial, and that's R in a nutshell.

Its still interpreted, not compiled you absolute brainlet. You cannot compare the two

Second this. R is amazing if all you work with are lists and matrices.

I use Excel at work. And R when they allow me.

>ugly trailing backslashes
>20 times slower than C
>lacking intellisense
>messed up variable scoping
>documentation far worse than c# and js

it sucks, too bad there isn't anything better than python anyway

give it another couple years, then yes

s/matlab/julia

>racket/scheme/lisp,
impractical meme shit.
>matlab, and any CAS is Veeky Forums
proprietary, expensive, and slow.

The problems with C++:
- Bloated with useless/overall harmful features
- They have to be backward compatible with retarded stuff
- You will make lots of errors, the code will be buggy
- Again, you WILL make buggy software and you will spend a lot of time fixing those bugs

Anyone who does not acknowledge these has only ever done home projects/only works alone on a couple thousand lines of code.

More like some weird drugs that mutated it into a big bulky monstrosity.

>proprietary
That's indeed shit and only reason Veeky Forums shouldn't use them.
>slow
That depends on what you're doing and how important your time is compared to the time running it. Usually your time writing it worths more.
>expensive
It isn't like you pay them yourself. Your institution does. And if you're a student they're dirty cheap anyway. Matlab is like $70 and Mathematica ~150$. That's one coffee/month for the >4 years you will use them.

>impractical
go on

humans are just bad at reading s-expressions.

alternatives have proven more useful.

additionally, the metaprogramming capabilities lisps give you just turn out to not be that useful in practice.

>nobody uses lisp

>brainlets are bad at reading sexp
so?
>metaprogramming capabilities lisps give users are poorly utilized by brainlets
so?

its harder to write, harder to debug

Sure, if you want to show off on a mongolian basket weaving forum, then LISP is a top choice.

choosing a programming language is about choosing the right tool for the job. C was designed to write an operating system. Lisp was designed for old school symbolic AI.

There are use cases for LISP, but as a general purpose language its just silly.

>- They have to be backward compatible with retarded stuff

This is the biggest downside of C++

all of the legacy bullshit that they can't change because it would cost billions of dollars to the global tech industry.

C++ could have been the undisputable go to general purpose language if it was forked. but now its just a weird frankenstein that lets you make preventable mistakes

Lisp was not designed for old-school symbolic AI, it was designed to be a better lambda calculus.

>it is harder to write
How so?
>harder to debug
This is actually false. It's a lot easier to debug due to its near homoiconicity.

>choosing a programming language is about choosing the right tool for the job
I disagree. Creating a programming language to solve your problem in is choosing the right tool for the job. What you're thinking of is re-solving solved problems. In that case you choose the same tool everyone else chooses. Modern lisps are programmable programming languages. You can write your own syntax. If C++ is a swiss army knife, then modern lisps are CNC mills.

Everyone who learns to program can sympathize with "I wish I could tell my computer to do something." Everyone who learns to program in a lisp can sympathize with "I wish I could tell my compiler to do something."

>- They have to be backward compatible with retarded stuff
Such as?
>- Bloated with useless/overall harmful features
Stop being autistic
>- You will make lots of errors, the code will be buggy
>- Again, you WILL make buggy software and you will spend a lot of time fixing those bugs
Garbage in, garbage out. Don't blame the language.

Imagine unironically being this stupid.

Ah yes. I remember when I was a college student...

You keep listing properties of LISP that separate it from other languages. What you aren't doing is showing that any of these unique properties are actually useful, practical, and improve productivity.

The reality is most programming is simple. We got to the fucking moon with hand written assembly.

People keep writing new languages all the time. Why is that? Why didn't we stop at, say, Fortran, COBOL, or C? Once you admit that syntax matters you've already made the entire argument for metaprogramming and it's literally just your own ignorance and biases preventing you from reaching the logical conclusion. Maybe you just think it should be left to others, but you're objectively wrong if you think it's just for college kids or memesters.

>Maybe you just think it should be left to others, but you're objectively wrong if you think it's just for college kids or memesters
>tips fedora

>The reality is most programming is simple.
Compared to what? Programming is a pretty legitimate candidate for one of the least simple things you could do e.g. that Douglas Crawford quote:
>Computer programs are the most complex things that humans make.
Also RE: Moon landing and assembly, you use low level programming like assembly or C for any embedded systems because the performance and fault tolerance are a lot more important when you're dealing more directly with hardware like that.
I think you're trying to make the argument that because the moon landing was a very difficult and impressive project and because assembly was used for it that this means assembly's good enough for any other applications. And I think where this argument goes wrong is that which tools work best for which situations isn't just a unidimensional scale from easier problems to harder problems. There are situations which are both not as impressive as the moon landing but also wouldn't be very good contexts to apply assembly to e.g. I seriously would not want to do web development in assembly.

>Also RE: Moon landing and assembly, you use low level programming like assembly or C for any embedded systems because the performance and fault tolerance are a lot more important when you're dealing more directly with hardware like that.
related: flownet.com/gat/jpl-lisp.html

>The only thing left was to get approval from the VP of engineering. The conversation went something like this:
>Me: I'd like to talk to you about something...
>Him: Let me guess - you want to use Smalltalk.
>Me: Er, no...
>Him: Lisp?
>Me: Right.
>Him: No way.
lol, executives literally keep a list of meme languages to avoid and Lisp is #2.

>I think you're trying to make the argument
Don't strawman me. Just ignoring this entire post.

Keep memeing for LISP. Read those Paul Graham masturbatory blog posts. Fiddle around in emacs all you want. The reality is very smart people have tried to make LISP work in the real world, and it just doesnt. There is a reason no serious company runs production systems with LISP

You're confusing me with the LISP user.
I don't care about LISP. Read my post you responded to just now again as its own thing because I think you completely misinterpreted it over your rage at that LISPer.

The google example is hilarious because they memed on python for a while, then ended up writing their own language. So they exhibit the same anti-pattern as everyone else that stubbornly refuses to embrace lisps.

>should we use lisp
>no it's too complicated let's write our own language from scratch
>should we prototype it in lisp first
>NO THAT'S A MEME
kek

Also how the fuck is that a strawman when I went out of my way to qualify my impression of what you wrote as just what I think you were trying to argue? That's the opposite of a strawman, I was making an honest effort to understand what you were trying to say without assuming I was getting it right.

>do matlab heavily and C/C++ for coursework and projects
matlab is literally what you give to the mech eng monkeys who think typing more than 5 lines of code is software development.
that being said, Simulink is the shit

moreisdifferent.com/2015/07/16/why-physicsts-still-use-fortran/

I studied Biomedical Engineering with a focus in computational modeling and we ALWAYS used Matlab in both undergrad and grad at 2 different schools. Yet none of the job postings I look at want Matlab. Should I learn another language? R and Python seem to be requested frequently

>tfw want to learn python and R
>tfw don't know any lesson plans to learn what I need to know to be ready to actually use it professionally.

What do lads?

Doubt it bros.
>import numpy as np
ls this what CS students are actually like?
Python is actually easier and more intuitive to use than R for manipulating data, although R basics are faster to learn.

Scala

>not mapping and hardcoding the environment as you go along

...

>t. neet
Computer time is cheap, programmer time is expensive.

Are you already a competent programmer? If so, learn the basics of Python data types and syntax, then choose some smallish projects to implement. When things don't work or you don't know what to do, look it up.

Can someone honestly disprove 's point? I really don't know why people consider Python as 'slow' when it saves so much development time? Sure, it's slow if you want to crunch numbers but for 80% of use cases I just don't understand why would you label it as 'slow'.

It's like saying that a car is slow in traversing 2000km and a plane is faster; sure, the car is slow but you can use it everyday whereas you barely use a plane unless you're traveling really long distances. Different tools for different jobs as they say, I believe.

Where does one find small projects?

I've literally gone to youtubes watched some lectures on python and did the exercise (converting DNA sequences and stuff to RNA) and other things like that. I didn't learn anything.

>Can someone honestly disprove 's point?
Literally no body can. You have to remember Veeky Forums is 99% filthy undergrad plebeians, they have no idea what the real world is like.

>Sure, it's slow if you want to crunch numbers
It's not though. Crappy Python is slow, but the solution to that is to not write crappy Python.
If you're doing large amounts of number crunching there's libraries designed for it rather than handling each element individually.

Depends on the "scientist". For my work I coded a solver in python that's 9x faster than it was in MATLAB that's already 10x faster than comparable finite element codes. I don't need to manipulate and analyze it til after I solve it. Since I'm already in python I do that in python too, not about to switch to R.

>Can someone honestly disprove 's point?
Computer time isn't "cheaper" than development time because you only develop your application once (in a loose sense, please don't start a nitpick semantics argument about versioning), whereas it will probably run a lot more than once when your consumers actually make use of it. That's kind of the whole point of programming and automation, you put in some work up front and get a reusable tool out of it afterwards instead of not putting in some work up front but having to do tasks manually over and over again.
i.e. It's an investment. Investments aren't immediately profitable, you give up something valuable up front with the expectation you'll reap a greater amount in benefits later.

I don't think most people on Veeky Forums are developing for consumers though. Most people here are probably just analyzing their own data.