/sci/ - Science & Math

>no Excel

>>9417423
/sci/ i need to learn c++/to program in two months dafuq do i do?

>>9417427
http://4chan-science.wikia.com/wiki/Computer_Science_and_Engineering#Basic_Programming_.26_Data_Structures

>>9417435
>http://4chan-science.wikia.com/wiki/Computer_Science_and_Engineering#Basic_Programming_.26_Data_Structures
thank you wise anon

>>9417423
>/sci/ related programming
>Keep insubstantial language wars to a minimum
>puts every fucking language that exist in the picture

>>9417423
/MatSci/ anon here
Wanna start programming as a hobby. I learned some very basic c++ stuff in HS and somewhat enjoyed it even though we were just solving some relatively easy mathematical problems. Which language should I start with? Continue with C-->c++ or just start from the beginning with python/java?

>>9417482
1 Java is horrible and not worth learning unless you really need to.
2 There's no need to learn C before C++. Stop listing to /g/
https://www.youtube.com/watch?v=YnWhqhNdYyk

>>9417482
It doesn't really matter.

>>9417505
>listing
*listening

>>9417505
I know jack shit about programming languages and it's just I've read on a couple of chans anons claiming that it's best to start with some c-like subset of java or something.
>>9417509
Fuck it python seems the most attractive one out there.

>>9417482
Speaking only from experience, I'm not a CS major, I don't think starting with C++ is too bad. Just make sure you start with a modern book that teaches you modern, standard C++11 (and newer) concepts. The 4th edition of the C++ book by Bjarne is as comprehensive as it gets, but it will be a bit too dense if you're rusty.
Python is decent if you don't care about performance, but managing the packages and dealing with two versions of python drives me up the fucking wall. I don't want to spend an entire day trying to get an environment to work, it's one of the reasons why I hate CS, that time wasted fucking around on the terminal trying to get shit to work. So, take my advise if you're going to learn python, and get Spyder with the Anaconda Navigator. It just werks without having to do anything.
I wouldn't suggest Java for science, if you want more specialized things then maybe try Matlab, R or Julia.

>>9417519
>Fuck it python seems the most attractive one out there.
Go for it

>>9417519
Just don't fall for the CS freshman/code monkey trap of constantly switching between languages and endlessly wondering if the one you're on right now is the PERFECT one for you while never making it beyond fizzbuzz with any of them. That's what people who like the idea of programming more than actually programming do, and it's pathetic (see: /g/).

Now pick some subject matter or projects that you're interested in or find a decent "scientific computing with Python" textbook and just do the work.

>>9417521
Friend of mine studying CS recommends pycharm 3.2 so I guess I'll have him guide me in the 101. C++ looks terryfing desu
>>9417524
Thanks for the help, anons.

>>9417533
If you're alright with it just go for it, but personally I wouldn't want to manually deal with packages ever again. Spyder is the popular choice for non-CS scientists for a reason.
C++ is a strange thing, the code looks impenetrable at first but when you get used to it the language can hook you.

>>9417519
>some c-like subset of java or something
There is no c-like subset of java.
>>9417533
>Friend of mine studying CS recommends pycharm 3.2 so I guess I'll have him guide me in the 101. C++ looks terryfing desu
But you said you already learned it, no?

C++ isn't scary once you realize you don't have to learn all of it.

>>9417553
I just learned the basics till loops and functions and it's been a while since then.

What kind of computers do you guys use?

I have a good laptop. 8gb RAM, i7 processor but every program i write that isnt a super simple 1styear physics student-tier program takes a while to compile. Besides cleaner code, any tips to speed it up?

What is the best python editor/compiler?

Best c++ editor? Tried eclipse but its a hassle to setup

Also can anyone point me to a starting point to learn parallel programming? In c++ probably. Fortran is kill isnt it? Anyway, I want to learn cluster computing.

>>9417423
i'm a bitcoin core developer.

>>9417592
>any tips to speed it up?
https://www.incredibuild.com/

>What is the best python editor/compiler?
Like I said above, I personally like Spyder. As far as compiling I don't see any reason to compile python code, it won't run any faster.

>Best c++ editor
Personally I like Visual Studio with the VAssistX plugin, which turns VS from crap to God tier. With IncrediBuild I can compile changes to huge programs in seconds (obviously after the first-time big compile)
Some old school programmers will use emacs or vim, not me.
I use XCode on my laptop but I haven't tried it on big projects.

>Also can anyone point me to a starting point to learn parallel programming?
Not sure but I would look for books on Amazon and check the reviews

>>9417607
Thanks for the info

>>9417592
>Also can anyone point me to a starting point to learn parallel programming? In c++ probably

C++ Concurrency in Action: Practical Multithreading by Williams

from
http://4chan-science.wikia.com/wiki/Computer_Science_and_Engineering#Parallel_Programming

Pytorch

for the love of all fuck please someone tell me a lighter tex editor than texstudio other than autistic vim

>>9417695
sublime with tex plugins

What are some neat physics projects in C++?

>>9417423

I've started studying Coq in my downtime.
Existing skills are in C, Erlang and Clojure. I'm liking the principle of what Coq offers but man, it takes a while to produce anything useful.

>>9417695
Notepad++

Is Julia a meme?

>>9417423
>Prgramming
Fix your shit, OP.

>>9417695
Gummi
gedit with latex plugin

Python + Pycharm is the best way to be productive. You're making tools that will only run few times. Prioritize your dev time over application performance.

>>9417423
>>9417433
What's wrong with Haskell?

>>9417695
Geany

>>9417505
>Java is horrible and not worth learning
Unless you want a job.

>>9417423
Im using python.
Image processing and measurements of computed tomography data regular images, processing of triangle meshes, measurements on topology data lately.

>>9417722
Finite element solver to visualize the heat equation.

>>9417695

I use emacs and love it

>>9417725

Coq is great fun to use, especially in emacs

>>9417423

Why System Verilog and not regular Verilog?

>>9418064
yes

>Documents prettier and faster than MS Word
TeXmacs
>Mathematical software easier than MATLAB
Maxima
>Modeling like Simulink
OpenModelica
>Easy prototyping and scientific programming
Perl Data Language
>Statistical package like SPSS
PSPP

By the way, TeXmacs has Maxima CAS integration. You should try them sometime.

>>9418359
>TeXmacs has Maxima CAS integration
Sounds very un-UNIX

>>9417423
Not sure if this is /sci/ or /g/ material, but I'm trying to make a neural-network play 2048. I already successfully did so with tic-tac-toe, but that was quite straightforward: 27 input neurons (3 bits for every square, O has it, X has it, or nobody has it, make it 1 if true and 0 if false) and 9 output neurons: Which square to pick next. Training was also quite straightforward; use a policy-gradient-ish way of learning where you log all the moves for a game. For the winner you backpropagate that the chosen output should've been one, for the loser it should've been zero. The earlier the move, the lower the gain.

For 2048 it's fucking awful. The inputs aren't straightforward: Sure you can just give the value of every square but that really doesn't work at all. You'd like inputs to scale between 0 and 1, and same for the outputs (the way I've built it, you could also go for between -1 and 1). So what I do is I divide the inputs by the highest one. The outputs are simple: up, down, left, right. Highest output determines the move. Learning is also really not straightforward, except that you don't want it to make moves that don't do anything, so those are backpropagated as 'that move should've gotten score 0', but everything else, I have no clue. What I'm trying now is the same idea as with the tic-tac-toe game: log the outputs, change the chosen output to a value, and backpropagate. The value to be changed to is not that clear. You'd prefer to have it between 0 and 1, so I created a function which rates a score of 0 at 0 and a higher score higher, but with an asymptote at 1. The result is really fucking shitty; it's as if it's not learning anything. In my early attempts at tic-tac-toe (Q-learning instead of policy-gradient) it at least went from a winrate of 0.4 to a winrate of 0.7 (against random moves); and when I perfected it the winrate was 0.995. But with 2048 just nothing improves at all.

>>9417505
All programming languagues are basically fine. Java has its strengths and weaknesses. No languague wars, please.

>>9418465
Sounds like a good use-case for empowerment.

http://ieeexplore.ieee.org/document/1554676/
https://arxiv.org/abs/1710.05101

You can model the goal of 2048 as keeping your future options as open as possible, or maximizing the channel capacity of between action space and sensory inputs over some time horizon.

>>9417427
Took me a week to learn it for my job. Just start writing shit in C++. It's got a lot of shit in its standard library premade for you.

>>9417423
>being such a brainlet that you don't know what Ocaml is

>>9417695
https://github.com/LandonPowell/zext
>implying you're intelligent enough for Zext

>>9418070
You won't find a job in science with it.
>on /sci/
>wants a job transcribing someone else's thoughts into code

>>9418518
Thanks for the links; I read the first paper but I don't really get any concrete ideas from it; the second one seems better in that respect but I'll have to look at it on less alcohol later. The idea of keeping future options open is an interesting one, I'll try that out; maybe I can use a simple heuristic like amount of empty squares, combined with amount of moves that actually do anything; something like that.

I feel that the way to provide input is also a big factor though; normalizing the values seems like a bad way to do it, but I'm not sure what else to use.

>>9418377
Rob, it's not the 70's anymore.

>>9418359
>tells people PSPP can substitute SPSS

kys yourself my man

>>9417505
Whats wrong with java, exactly?

>>9417970
yes

>>9419093
Yea Java isn’t that bad desu

>>9418359
Maxima is CAS, not numerical computation software imbecile.

>>9418617
>You won't find a job in science with it.
You likely wont find a job in science. You might not like that thought but it's the truth. In which case you'll be looking for the next best thing, that'll likely be some form of software dev position. So you'll almost certainly be using Java or something else, I'd advise you (and anyone else reading this) to learn it, if for no other reason than as a backup should all else fail (which it will).

>>9419093
slow for numerics, too verbose for academics

>>9418060
Irrelevant in scientific computation. I guess the second one is the original and some fag edited out R for Haskell. R makes more sense with the rest.

>>9419232
>being this retarded

>>9419263
>recommends an alternative to a tools he has never used

>>9419279
>implying you know

>>9417970
It's promising but nowhere near stable enough.

>>9419279
>to a tools
learn to speak, cunt

>>9419312
Yeah, I was thinking about the other tools as well as anon I was replying to obviously has no idea what he is talking about.

>>9419318
You seem mad for some reason, maybe you shouldn't be posting on 4chan. Take a break.

>>9419332
>maybe you shouldn't be posting on 4chan
Once here, forever here.

>>9419238
So I only know Java and a bit of Python from uni, how do I get out of the Java trap? What’s the next best language?

>implying all computer algebra systems can't do number crunching
That poster should off himself.

This devolved into role-playing pretty quickly.

>>9419368
What the fuck are you talking about? You sound like a crazy person.

>>9419359
Depends on what you're doing. Python encompasses all but low-level stuff if you learn the right libraries. Symbolics (sympy), numerics (numpy), visualizations (matplotlib), stats (pandas) and specialized modeling is possible.

>>9419375
>Python encompasses all but low-level stuff if you learn the right libraries
Isn't that almost every programming language?

>>9419382
Not every language has those libraries you need. And those that do (certainly Java, C) aren't as easy to use as Python. it would be great to use Ada for everything but it isn't feasible.

>>9419398
Didn't Fortran have more libraries and also better suited and more optimized for scientific programming?

>>9418481
Java is just a more horrible version of C# with similar portability problems. It has 0 redeeming features.

>>9419093

>>9419416
>is just a more horrible version of C#
>similar portability problems
I don't expect high quality replies, but didn't expect lies either.

>>9419404
Fortran is good choice for numerical crunching such as in large-scale simulations. Also it is simpler to program in (no pointers and such) and has built-in matrices handling. Most big libraries target both Fortran and C/C++. In physics it is still in widespread usage (especially prevalent in meteorology and climate modeling) but so is C/C++. (Languages that are arguably usable in more cases.)

>>9419421
>Java is slow meme
Compared to what? C? Yes, it is slower. But it is faster than Lisps, Node and similar to Go.
>Java is insecure meme
The long dead browser plugin was insecure.

>What are you programming scientifically
I tried working on this idea I had a while back of a way to localize your position on a map of landmarks given noisy measurements of the landmarks in your immediate vicinity with the idea that it might be a good way of figuring out your position and orientation if you somehow find yourself lost in space with nothing but a database of stars and the estimated location of the stars around you.

I got as far as modelling the noisy measurements from the ship which are these rotated normal distributions with a variance in the direction of your ship that gets bigger the farther away the star is. My thought was that you'd be able to localize the transverse position of the stars pretty well just by looking at them but your estimate of the depth would be limited by the instruments you had on board.

With that and a model of the database as a bunch of normal distributions (since those measurements are known only to some finite precision) the way I figured you'd compare the current estimate of your position and orientation to another candidate estimate is by multiplying the estimate distributions with the databases' and integrating over the plane (working on the 2D case first) and just seeing which is bigger, but now I need a way of choosing candidates.

>>9419433
>t. has never tried running java from windows in unix or android

The JVMs are garbage.

>>9417423
>no R
>no Excel
>no GIS software

>>9419685
>Excel

>>>/biz/ness monkeys are that way

>>9419722
So you went as far as replacing C++ in that picture for Java? It must be personal.

>>9419722
>Works the way a computer really works
Role-player

I like the JVM but not necessarily Java, plan on learning Clojure (and Clojurescript). Bytecode is greatly under appreciated.

https://www.wolframalpha.com/input/?i=Z%2Fsqrt(x)

https://www.wolframalpha.com/input/?i=PRIME%5E-1

https://www.wolframalpha.com/input/?i=(sqrt(2)%5E-1)%5En

I don't get it really.

Do you guys just want the fastest way to generate primes, or to check primes?

What do you want PRIME numbers for?

P.R.I.M.E. = Personally? Really, I Much Enjoy?

Primarily Return Integers Multiplication Emissions (By Back Propagation)
//
我不明白這一點真的。

難道你們只是想生成素,或檢查素數的最快方法?

你要素數是什麼?

P.R.I.M.E. =個人? 說真的,我非常喜愛?

>主要返回整數乘法排放(<<速率的傳播<)

>>9419925
nobody can read your moonrunes

>>9419939
I can, but it reads like a retard wrote it.

>>9419939
Then My Name Is nobody At This Point.

https://www.wolframalpha.com/input/?i=sqrt(7)+sqrt(5)+sqrt(3)+sqrt(2)+sqrt(i)+sqrt(-1)

[math](-1)^(3/4) sqrt(210)[/math]

>>9419944
Then describe to me how you want this maths done. I have no difficulty in crafting it, it's just YOU guys have a hard time agreeing that the new world order is being crafted by some random shit-poster.

If anyone was gonna call the cops on me it would've been 'months' ago.

Align your fucking time streams with mine.

>>9419944

He just wants to be the next bacon rider

Ignore him

>>9420030
他只是想成為下一個熏肉騎手

不理他

西蒙主宰刺激通信(S'#'DAC)
西蒙按摩時間(SMT)

Simon Dominates Stimulation Communications (S'#'DAC)
Simon Modulates Stimulation (SMS)

>>9419685
>no GIS software
GIS generally relies on other languages for programming though. Like QGIS with python and C++ (also R with some packages).
I sure hope you're not an ESRI brainlet

>>9420046
so what if I like having a GUI?

>>9417423
I'm personally obsessed with C++ and it is the only programming language you will ever need. There is literally nothing the language cannot do. You can implement anything you want. Any weird convenient language feature that exists in other languages you can create in C++ in your own custom way

My favorite style of programming is C++ template metaprogramming in C style. It is the most performant and flexible code you can get. I help build machine learning infrastructure at my job and we do it this way with excellent performance results and agility

>>9417423
i fucking hate perl but it is used a lot in informatics.

>>9420049
it's free nigga

What operating do you guys run?

Whats the best os for controlling clusters?

Why do cern/fermi etc all run fedora/rhel based distros?

>>9420167
>Why do cern/fermi etc all run fedora/rhel based distros?

They are a famous adopter of Ceph, which is developed on Redhat derivatives.

>>9420167
>>What operating do you guys run?
GNU/Linux, specifically Ubuntu. Only because that's one of the officially supported operating systems for ROS
>>Whats the best os for controlling clusters?
Is that even a question? GNU/LINUX by a fucking long shot. Fuck I'm not even sure just what job managers run on wangblows.
>>Why do cern/fermi etc all run fedora/rhel based distros?
And CentOS. Cause they're fucking solid and you can get great support for them. Also I would advise you to to encourage your admins not to install the updates which fix the intel bug. HPC stuff is gonna take a HUGE performance hit.

>>9420052
>You can implement anything you want. Any weird convenient language feature that exists in other languages you can create in C++ in your own custom way
This is most definitely not true. There are plenty of programming structures you cannot make properly in C++ at all, such as garbage collection, coroutines, or forms of namespaces that do not fit the C++ model. You can make hacky broken approximations, but nothing that really behaves like the real thing.

There ARE a lot of things you can make yourself in C++, it truly is powerful, and I love it for that. But it is NOT all-powerful in this regard, and we shouldn't praise it beyond what it can actually do.

>>9418587
I'm gonna learn this in spring semester, what am I in for?

>>9420192
First off, you can do GC in C++ it's just weird. Second off, GC is just a crutch for broken languages with garbage memory management models. RAII is the superior method and always will be because it completely eliminates all need for memory management.

You don't have to think about memory management at all with object scope lifetime and smart pointers.

Coroutines fair enough but there are libraries for them specifically boost.

SageMath is a fucking nightmare to use.
That is all.

>>9417505
>>9419093
>>9419421
>The absolute state of mathematicians
Java is one of the easiest to use, dumbed down, universally used, and *relatively* quick languages around. Short of jumping into C/C++/Rust not much beats it. If you think it's hard it's because you're an idiot, people who don't even speak English manage to learn it through its entirely english based documents.

>>9420215
>RAII is the superior method and always will be because it completely eliminates all need for memory management.
Imagine being this misinformed

>>9417482
I had a similar experience except did Java in high school and not C++. I started learning Python (would recommend for science stuff) and JavaScript (would recommend for web stuff but not science stuff). I've been slowly spending less time with JavaScript and even more time with Sage instead. I say just go ahead and start with Python and start exploring once you become more comfortable programming.
t. Materials Science / Electrical Engineering student

>>9420040
>>9419925
>>9419939
>>9419944
Why have there been so many Chinese posts on /sci/ recently? I'm not really complaining since they've all used Traditional Characters so far but most of the posts aren't written like any other Taiwanese person would write or speak in my opinion.

>>9417423
>What are you programming scientifically, /sci/?
Algorithms for solving large systems of Equations.

>>9417695
There isn't a good light weight tex editor with a nice to look dark theme, which works decently well and doesn't FUCK UP HALF YOUR SCREEN WITH GARBAGE.

For me, it's pic related.
The fact that it's easy should really not be overlooked. Super accessible for new learners who don't have to get bogged down in teaching themselves unintuitive syntax.

Now that I'm here, I want to write a pretty large and fast simulation in Python that uses parallel processing. Should I just use threading or should I convert my .py files to .exe to possibly get more CPU usage?

>>9420264
>should I convert my .py files to .exe to possibly get more CPU usage?
You mean you want to compile them?
That probably is a very good idea.

>>9420220
Explain

>>9420266
Why would compiling be better though? I imagine it'd be faster loading but that's it?

>>9420274
>Why would compiling be better though?
Because interpreting software comes with a lot of overhead, I mean it probably makes not *that much* difference since you are probably relying mostly on compiled libraries.
You should probably read up on the difference between interpreted and compiled code, there is a reason high performance code isn't written in python and interpreted in real time.

Try implementing Matrix multiplication with a for loop and interpreting that.
Your implementation will be orders of magnitude slower the LAPACK code, or just the same thing in C.

>>9420264
While Python is probably my second or third favorite language ever, it is very, very slow. Like VERY slow. Any sort of intense computation should NOT be written in Python, it's more of a hobbyist and prototyping tool. Small projects where speed isn't essential is where it shines.

>>9420264
>>9420266
You can't compile Python you mongoloids. All things like py2exe do is stick a Python interpreter with all the necessary libraries in a self-contained folder so people who don't have Python installed on their system can run it.

Dumb CS freshmen. This thread is ass.

>>9419434
Fortran is also still big in aerospace (modern and f77), and isn't going anywhere anytime soon.

>>9419718
confirmed for never having worked in industry. excel is the glue that holds everything else together.

>>9420574
Bro, this is the SCIENTIFC programming thread, not the INDUSTRY programming thread

>>9420597
>he works for free

>>9420574
>excel
fuck off, this is SCIENTIFIC programming. python is usually the shit holding the rest together.
>>>/out/

>>9420625

>>9420625
>>9420597
>there is no scientific industry

>>9420046
I mean wouldn't you say it's mostly SQL

[math]\color{#781b86}{\text{P}}\color{#591da1}{\text{r}}\color{#472cb7}{\text{o}}\color{#3f43c7}{\text{g}}\color{#3e5dcf}{\text{r}}\color{#4276cd}{\text{a}}\color{#498ac3}{\text{m}}\color{#529bb4}{\text{m}}\color{#5da8a2}{\text{i}}\color{#6bb18d}{\text{n}}\color{#7ab77a}{\text{g}}\ \color{#9ebd59}{\text{c}}\color{#b1bd4d}{\text{h}}\color{#c3b945}{\text{a}}\color{#d2b23f}{\text{l}}\color{#dea63b}{\text{l}}\color{#e59537}{\text{e}}\color{#e67d32}{\text{n}}\color{#e3602d}{\text{g}}\color{#de4027}{\text{e}}\color{#da2121}{\text{:}}[/math]

[math]\color{#781b86}{\text{W}}\color{#691c92}{\text{r}}\color{#5b1d9f}{\text{i}}\color{#4e1fab}{\text{t}}\color{#4829b4}{\text{e}}\ \color{#3f3ec5}{\text{a}}\ \color{#3e56ce}{\text{p}}\color{#3f62cf}{\text{r}}\color{#416dce}{\text{o}}\color{#4379cc}{\text{g}}\color{#4682c8}{\text{r}}\color{#4a8bc2}{\text{a}}\color{#4d94bc}{\text{m}}\ \color{#57a1ac}{\text{t}}\color{#5ca7a3}{\text{h}}\color{#62ab99}{\text{a}}\color{#69b090}{\text{t}}\ \color{#77b67d}{\text{m}}\color{#7fb874}{\text{a}}\color{#87ba6c}{\text{k}}\color{#8fbc65}{\text{e}}\color{#98bd5d}{\text{s}}\ \color{#a9bd52}{\text{/}}\color{#b2bd4c}{\text{s}}\color{#babb49}{\text{c}}\color{#c3b945}{\text{i}}\color{#cbb742}{\text{/}}\ \color{#d7ae3d}{\text{r}}\color{#dda93b}{\text{a}}\color{#e0a13a}{\text{i}}\color{#e39938}{\text{n}}\color{#e58f36}{\text{b}}\color{#e58334}{\text{o}}\color{#e67831}{\text{w}}\ \color{#e25c2c}{\text{t}}\color{#e04d29}{\text{e}}\color{#de3e26}{\text{x}}\color{#dc3024}{\text{t}}\color{#da2121}{\text{.}}[/math]

>>9420218
Underrated post.

>>9419232
Macsyma (aka Maxima) can do numerical analysis too.

Just use Ruby, it's insanely simple, or even Haskell, it's syntax is similar to math notation and is purely functional. Great for math work.

>>9420665

hue of character i = hue range * i / length of sentence

>>9420731
Oh come on, all of the supposed programming experts ITT and this is the best that you can manage?

Wrote a program for this guy >>9419770

>>9420847

that's the program. hire some pajeet to do the codemonkeying.

>>9420862
t. Ideaguy

>>9420862
Based role-player.

>>9420873
>>9420881

it's busywork. you'd just have to order the html colors that you wanted by hue and then use the formula.

>>9420857
What program/language is that?

i can see exactly how to do the program and so it is boring.

>>9420888
The one and only.

>>9420891
yeah programming is so boring let's just spew some more platitudes about this or that programming language

JAVA SUX
HASKELL IS MATH
XDDD

>>9419093
there are only two things wrong with java. the initial design and the implementation.

>>9420196
A somehow perfect blend of functional and procedural programming. Jerking it to your own code. Thinking you're a super-genius. Hearing the voice of god.

>>9420665
Shade text horizontally using rainbow gradient.
>>9420731
Kek, idiot.

>>9420531
You can compile Python with a Python compiler, brainlet.

>>9420264
You can make .py executable by adding #!/usr/bin/env python to the beginning of the code or compile and run it by running py_compile example_code.py in cmd prompt for Windows or chmod +x example_code.py followed by running ./example_code.py for a terminal for Linux.
>>9420531
You can, in fact, compile Python. While it doesn't need to be compiled in that it is an interpreted language, one may still compile its code quite easily.
For Python 2
https://docs.python.org/2/library/py_compile.html
For Python 3
https://docs.python.org/3/library/py_compile.html
Compiling Python source code can be a very good idea since the code may be able to run significantly faster, as >>9420266 even states.

>>9417423
I use R for all my analyses, modeling, and simulations.

>>9421211
>>9421234
>Compiled
>It still has be be INTERPRETED by an INTERPRETER
Fair enough, but people usually mean "compiled to native machine code" when they say "compiled".

>>9421093
>Shade text horizontally using rainbow gradient.
>Kek, idiot.
'Kek, idiot' yourself, role-player.

Cern uses windows

/sci/ blown the fuck out

>>9421686

>>9421686
OS autism is from /g/. It makes sense that a place where non-programmers work at uses windows. Anything is either for hipster autists, or autistic autists.

>>9421691
Anything else*

So I dont mean to start a war but this prob will

Im an undergrad physics student. Want to go into plasma physics specifically.

Should I learn c++ or fortran? The grad students tell me that fortran is kill but im not sure

>>9421686
>>9421690
https://youtu.be/RYsaGwqi7hQ?t=111
Stop being a nigger.

>>9421713
You should learn plasma physics.

>I want to go into art. Should I learn how to suck dick or how to drive trucks?

>>9421713
>it's another "what language should i learn" post

>>9421735
Why are you even shitposting on this board?

Different areas of physics use different languages. Plasma is relatively new so I dont know if theyd use fortran or not.

>>9421691
They dont, is literally one guy using windows in one computer that the other guy is bragging about.

>>9421713
Just learn C++17. You can pick up Fortran 2003/08 when/if you need it later.

>>9421744
I cannt advise you on physics specifically but I would listen to the grad students. In the end, FORTRAN and c++ interoperability is not bad and if you really have to use a FORTRAN library you can generally interface it from C++.

>>9421725
>hit enter a few times at the top
>then again in the middl
>"cd work", "ls", "cd grancher"
>one more complicated "wget" command (though not actually complicated, just more work to memorize)
>a few more empty lines at the bottom
LITERALLY Karlie Kloss tier

>>9421713
fortran is not at all kill (real answer is learn both)

>>9420231
>> 9420040(你)
>> 9419925(你)
>>9419939
>>9419944
为什么有过上/ SCI这么多中国的职位/最近还好吗? 我真的不抱怨,因为他们都使用繁体字,但迄今为止大部分职位都没有写的任何其他台湾人写信或在我看来说话。

>>9420665
编程挑战:编程挑战:

编写一个程序,使/ SCI /彩虹文本。

>>9420665
阴影文本使用水平彩虹渐变。
>>9420731
KEK,白痴。

>>9420665

字符I =色调范围的色调*句子的I /长度

>>9421775
>crazy_strings.rb
>.rb
>code

>>9417427
get real familiar with pointers and malloc and free (or "new" and "delete") early on.

>>9419093

>>9422169
That was already posted and the original image applied to C++. Essentially you can say the same for almost any object oriented language.

[math]\color{#781b86}{\text{P}}\color{#641c97}{\text{o}}\color{#501da8}{\text{s}}\color{#482ab5}{\text{t}}\color{#4037c2}{\text{i}}\color{#3f48c9}{\text{n}}\color{#3e58cf}{\text{g}}\ \color{#4378cd}{\text{m}}\color{#4785c6}{\text{i}}\color{#4c91bf}{\text{n}}\color{#529bb4}{\text{e}}\ \color{#61aa9c}{\text{s}}\color{#69b08f}{\text{i}}\color{#73b482}{\text{n}}\color{#7db876}{\text{c}}\color{#88ba6b}{\text{e}}\ \color{#a0bd58}{\text{n}}\color{#acbd50}{\text{o}}\ \color{#c3b945}{\text{o}}\color{#cdb541}{\text{n}}\color{#d5af3e}{\text{e}}\ \color{#e19d39}{\text{e}}\color{#e59136}{\text{l}}\color{#e68133}{\text{s}}\color{#e57130}{\text{e}}\ \color{#df4928}{\text{i}}\color{#dd3525}{\text{s}}\color{#da2121}{\text{.}}[/math]

[math]\color{#781b86}{\text{Has}}\color{#621c99}{\text{ a }}\color{#4e1fab}{\text{par}}\color{#462eb9}{\text{ame}}\color{#3f3ec5}{\text{ter}}\color{#3e50cc}{\text{ to}}\color{#3f62cf}{\text{ co}}\color{#4273cd}{\text{ntr}}\color{#4682c8}{\text{ol }}\color{#4b90bf}{\text{gro}}\color{#529bb4}{\text{upi}}\color{#5aa4a7}{\text{ng }}\color{#62ab99}{\text{so }}\color{#6cb28b}{\text{tha}}\color{#77b67d}{\text{t l}}\color{#82ba70}{\text{ong}}\color{#8fbc65}{\text{er }}\color{#9cbd5a}{\text{pos}}\color{#a9bd52}{\text{ts }}\color{#b6bc4a}{\text{don}}\color{#c3b945}{\text{'t }}\color{#ceb541}{\text{sur}}\color{#d7ae3d}{\text{pas}}\color{#dea53a}{\text{s t}}\color{#e39938}{\text{he }}\color{#e58935}{\text{cha}}\color{#e67831}{\text{rac}}\color{#e3632d}{\text{ter}}\color{#e04d29}{\text{ li}}\color{#dd3725}{\text{mit}}\color{#da2121}{\text{.}}[/math]

I took modelling & sim last semester and got an A-

I dont understand why I didnt get an A

What type of final project is A material?

I did a monte carlo simulation of plasma dynamics. I know for a fact it was accurate. It was 1200 lines in matlab. Was it too short or something?

Someone who got an A, what was your project?

>>9423027
Post it.

>>9423032
its on a usb drive 60 miles away. i wont have access to it until next monday. i'll ask again then with it

>>9423027
>Someone who got an A, what was your project?
Didn't anyone get an A in your class? Also it obviously depends on your professors

>>9417592
Do you mean multithreathing or map-reduce?

Hadoop (Open Source implementation of Map-Reduce) is the thing to learn for multiple computers working together.

>>9420167
Windows personal computer, the university manages the cluster (GNU/linux based).

>>9420898
Do you like wolfram and is it only good for maths?

>>9423363
>Do you like wolfram
Yes.

>is it only good for maths?
No.

My favorite thing about WL is the simplicity of the programming model. Literally everything in the language is an "expression", or a tree with nodes of a certain symbolic flavor. You can take anything in the language and use FullForm on it and you'll get a 1D representation of this tree (TreeForm for a 2D graphical representation) that is a completely valid piece of Wolfram language code. There's no syntax for, say, a for-loop that goes beyond this framework like in other languages. This allows for basically unlimited meta-programming and an extremely expressive pattern matching system (since everything is so structured). Fancy formatting and images are essentially just syntactic sugar on these trees granted by the notebook interface.

Pic is an example. Not only is this program I wrote with crazy typesetting just another expression (I could have entered that instead and gotten the same result), but the graphical output it produces is, again, just an expression.

Takes a while to grasp the significance of this but when you do it's hard to switch back to anything else.

>>9417592
Your i7 is probably shit if it's a laptop one. And you're likely not programming for multiple cores so it would be faster if you had a faster per core i5 or i3 anyway.

Python editor:
Sublime or some very simple text editor. Python isn't compiled.

For compiled languages vim or emacs are best

Fortran isn't kill at all. Matlab is super easy to write programs that are ran in parallel but is a shit language.

>>9423558
>Your i7 is probably shit if it's a laptop one.
How to buy a good one if I want a laptop?

>>9423558
This is good advice.

>>9423520
I like this a lot I think I came a little

>>9423558
>For compiled languages vim or emacs are best

Fuck off >>>/g/

>>9423790
>>9423792
Fuck off matlab shill.

>>9423845

>>9422898
Thats pretty gay son.
A cool kind of gay.

>>9421519
ffs brainlet there are several tools for compiling Python code to native machine code fucking google it dumbass

>>9423845
>using autistic software to fit in with /g/

>>9424221
There are not.

>>9424246
>being this uninformed
>what is Nuitka

>>9417423
>trashkell
sorry I thought this thread was for scientific (read USEFUL) programming languages?
https://www.youtube.com/watch?v=iSmkqocn0oQ

>>9423520
Isn't WL totally paid?

>>9420052
Sure, you can spend a weekend, or a week, trying to properly implement some abstract concept like neural networks.
Or fire up Python, type "import libraryname" and call it a day.
You can spend half an hour writing a program that will process data from a .csv file. Or write an awk script in 30 seconds.
You can spend an hour or two implementing regex in your program or learning a new library for it, or process the logs you want with bash + sed in a few minutes.
tldr: there's this thing called "right tool for the right job".

>>9424347
It costs money if that's what you mean.

>>9417423
Joy

>>9424369
>you need new programming languages to do simple library tasks

Go back to the >>>/g/hetto

>>9424369
>some abstract concept like neural networks.
Freshman.

Compuational chemist here,

I'm just about finished with a RPMD (Ring Polymer Molecuar Dynamics, its closely related to Feynman Path Integrals) code that interfaces with Gaussian for its electronic structure gradients.

>>9424228
What does autistic means to you? You are beyond confused to think matlab is the only tool.

>>9425070
Looks like Maxima to be honest.

>>9425070
>not using newton's method

>>9417695
Texworks for Win, Linux
Texstudio for Mac

>>9424369
Fucking this.

Parsing a huge-ass CSV file is a single line in Python/Pandas. Want to try out some k-means clustering on your data? Takes about a minute with Scipy.

Unless you need critical raw performance, I don't see any advantage in using C or any other compiled language. It's just going to take you days rather than minutes fiddling around with datatypes of some obscure library that you obviously need to correctly link into your project too.

pip install --user fuckthisshit

>>9419242
I came into this thread to bash OP for not including R in the image, because of how ubiquitous it is in any kind of informatics based field, especially bioinformatics.

>>9425468
Newton's is a great way of making really complicated symbolic expressions

[eqn]
-\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}+\frac{\pi }{4}-\frac{\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\cos \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)}{\sin \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)-1}-\frac{\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\cos \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\frac{\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\cos \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)}{\sin \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)-1}\right)+\frac{\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\cos \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)}{\sin \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)-1}}{\sin \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\frac{\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}+\cos \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)}{\sin \left(\frac{\frac{1}{\sqrt{2}}-\frac{\pi }{4}}{-1-\frac{1}{\sqrt{2}}}-\frac{\pi }{4}\right)-1}\right)-1}
[/eqn]

>>9417427
do you need the OOP part?If yes, then you're lost my friend. C++ is a mess, just to give you an idea, you can actually enable or disable inheritance in any method using "virtual", after you pass the clusterfuck within vtables, well, then you're going to learn a bootleg version of metaprogramming in C++ using templates, but I doubt you will be able to get to this point, hell, I think you will not even gonna be able to differentiate malloc/free from new/delete

>>9417427
lol, I forgot the boost library, I'm smelling your asshole bleed right now.

>>9417458
What about Julia or LISP?????
GTFO brainlet!!!!1111

>>9419685
you're confusing /g for /sci buddy, gtfo spreadsheet jockey.

What do you guys think about Prolog?

>tfw the german permavirgin who maintains a dependency you use fucking broke everything and you can't do work

>>9425681
>you can actually enable or disable inheritance in any method using "virtual",
>differentiate malloc/free from new/delete

/g/, stop pretending you know anything and go back to your containment board.

>>9425702

>>9417423

You only need C++, anything else will just make a worse version of whatever you are doing

>>9417695
>not using vim
sorry friend, but you're already too autistic to be saved

>>9417435
>boost library
>>9418583
>>9422159
>>9425681
>>9425683
LOL niggas I don't nedd it for work, i'm a mathlete that was invited to an ICPC team.

>>9419242
haskell has been used for automatic provers, AI, and category theory research

>>9426505
What about Prolog?

>>9426505
>AI
lol

>>9421686
Why do wage cuck scientists dress like they just left the gym or got out of bed?

>>9426565
Why is /sci/ so retarded they think AI is /x/-tier despite the fact we have several prototypes that act extremely similiar to the human brain.

I would give you the benefit of the doubt and assume you work with "AI" in which case you are laughing because we tend to refer to it as GI or just Intelligence; but I'm pretty sure you're just a faggot.

>>9425702
>he's never programmed in R

You need to return to >>>/g where other script kiddies and wannabe's hang out

>>9426663
>we have several prototypes that act extremely similiar to the human brain.
lol

>>9426747
>lol
I miss roflmao.

>>9426663

Going to be learning LabView for my internship in a week, what am I in for?

>>9426773
It's amusing how often that image is posted, despite there being nothing in that book arguing against the claims in that image. I can only imagine that nobody who posted that image actually read the book.

>>9426808
The point is to actually read it and see what "AI" actually is in practice. Stop refusing to learn beyond popsci bullshit.

>>9426843
But what AI actually is in practice today, has very little bearing on what it may become in the future. And it's not like that book argues for any apparent limitations on what the future might hold in this area. Trying to estimate future development is fundamentally not something you can do just by shitting on everything not possible with current technology.

>>9421686
And Comic Sans

>>9422898
There was something like that on /dpt/.
Are all Mathematica users rainbow faggots?

>>9426851
>But what physics actually is in practice today, has very little bearing on what it may become in the future. Trying to estimate future development is fundamentally not something you can do just by shitting on everything not possible with current technology.

>>9427052
Physics is a study of how the world works, not an engineering principle of trying to accomplish something. If present-day AI has claims to make about theoretical limitations of what an AI might fundamentally do, then that part does have something to say about where the future might go. But it doesn't, so the analogy doesn't hold water.

>>9427045
No I just like to rewrite old code sometimes.

>>9426795
Pain.

Wrote a Stan program for multivariate polynomial regression.

>>9420189
No, Richard, it's 'Linux', not 'GNU/Linux'. The most important contributions that the FSF made to Linux were the creation of the GPL and the GCC compiler. Those are fine and inspired products. GCC is a monumental achievement and has earned you, RMS, and the Free Software Foundation countless kudos and much appreciation.

Following are some reasons for you to mull over, including some already answered in your FAQ.

One guy, Linus Torvalds, used GCC to make his operating system (yes, Linux is an OS -- more on this later). He named it 'Linux' with a little help from his friends. Why doesn't he call it GNU/Linux? Because he wrote it, with more help from his friends, not you. You named your stuff, I named my stuff -- including the software I wrote using GCC -- and Linus named his stuff. The proper name is Linux because Linus Torvalds says so. Linus has spoken. Accept his authority. To do otherwise is to become a nag. You don't want to be known as a nag, do you? Be grateful for your abilities and your incredible success and your considerable fame. Continue to use that success and fame for good, not evil. Also, be especially grateful for Linux' huge contribution to that success. You, RMS, the Free Software Foundation, and GNU software have reached their current high profiles largely on the back of Linux. You have changed the world. Now, go forth and don't be a nag.

Thanks for listening.

>>9428629
>GPL and the GCC compiler. Those are fine and inspired products

Fuck no.

>>9426505
>category theory research
Absolute bullshit, saying this as someone who is an aspirant (or whatever post-major is called in the West) writing papers in homological algebra and categories.

Are there any other languages besides pic that can keep track of units of measure?

>>9429424
http://reference.wolfram.com/language/guide/Units.html

Wolfram does but I've noticed it can really slow down computations. Does F# have benchmarks to compare with?

>>9417423
could someone explain it to me, why is python or C++ considered as a /sci/-related programming language, and i.e. Java isn't?

I plan to study C++ and Mathematica this year
Is this a mistake?

>>9429865
seplesples truly is

>>9429736
Python and C++ are used in scientific/numeric computing. Java isn't.

anyone work with mass spec data?

>>9421521
>He doesn't know LaTeX has text shading
I D I O T

>>9421775
>a wget with a link is "more work to memorize"
>he thought the output of the wget command was fucking flags
You're not funny, just stupid. Relocate to /pol/.

>>9426422
>too stupid to know how to learn C++ in a huge timespan
>is a "mathlete" aka a turbo-virgin
>admits this on the internet

>>9429876
>>>/g/, you don't belong here. Leave and don't come back.

>>9429446
I have found none yet. Seems like a gimmicky feature, but would be something that would be nice to use occasionally.

>>9417521
Spyder ftw. I've used Sublime(required for a class) and Rodeo, but Spyder is the single greatest IDE I've ever used.

>>9430378
>>too stupid to know how to learn C++ in a huge timespan
I'm not too stupid, I just want too now a proven way of learning from people who have already; so I don't go about mindlessly learnign shit I will never use.
>>is a "mathlete" aka a turbo-virgin
fair point

>>9429736
Custom. Python is not that good of a language but it got trendy somehow and now everybody repeats the same thing. You would probably get better results by using Perl Data Language.

>>9419685
>Excel
Let me guess, you're a "manager"?
>>9420574
Nobody doing actual work in industry uses excel. Only terminal brainlets who got a shitty manager job with daddy's contacts use excel.

>>9424228
>2017+1
>not using vim or emacs
Confirmed for brainlet.

>>9426851
Terminal brainlet desperate to hold on to his AI fairytales detected. I feel sad for you anon, but at the same time you disgust me.

>>9432072
t. unempoyed

>>9417433
reeeee i should be learning how to use r instead of posting here god fucking damn

>>9417725
Read Chlipala's book and get familiar with the results from Voevodsky's group (especially if you have a background in math). Coq is greatly underappreciated, it is an amazing tool that very few people utilize.

>>9420312
you should learn to use generators, look at the package itertools

>>9432329
I believe that all those packages and libraries are trying to fix a language that is broken by essence. Python is of no use for me.

>>9432337
generators are built in (using the yield command) but itertools provides some other stuff. But I agree, sometimes python just doesn’t cut it.

How often do you guys comment in your code?

>>9432632
no one else is gonna look at it so NEVER

>>9432072
:^{ } (====8

>>9417423
Is Racket actually used for scientific/engineering work? Curious because it might be interesting to try out a modern Lisp dialect.

This is the only tip youre gonna need

Learn python first, it can accomplished anything you ever wanted to do faster than any language.
(even multiprocessing works somewhat ok)

When you've decided on the best way to implement, identify speed bottleneck and alleviate them by swapping parts to Numpy/Scipy, Cython, C/C++ extension

Only when all above fails, move to pure C/C++

>>9432632
>>9432640
I hardly ever write any code anymore, but if I do, I add concise, well-placed, and useful comments, that I usually cannot decipher nor reconstruct when I need them later

Anyone else writing this obsolete fucking shit?

No? Just me? Okay

>>9434168
The shit in question

>>9434151
>recommending python
Shit advise.

>>9434170
What dis?

>>9434170
My astro-nigga (Or are you a geophysicist or something?)
Y'all on 8.6 yet? My team's stuck with 8.5 for a couple legacy projects
Any good stuff?

>>9417423
The fact python is so used in science is more disgusting than the fact Fortran is. Heck, I feel more comfy about Fortran still being a language used a lot, at least that's what it was designed for. Python is a shitty scripting language which is outstaying its welcome.

>>9434744
There is nothing disgusting about using modern Fortran for scientific work. I could only understand some frustration if you had to actually edit or write a lot of F77 or earlier stuff. F90 or later is simple, fast, and straightforward and really ought to be used more often over all the trendy scripting languages.

>>9435153
Going to rescue this thread from the depths of the last few pages by backing this up
FORTRAN is nice and clean to use, I wouldn't write anything heavy in anything else
Maybe C, but idk

does anyone else run into issues with making multiple plots with a for loop in RStudio? I think I'm only getting the first or last plot in the plots pane.

>>9417427
if you know java or any other oop language you should be fine. make sure to go over pointers and malloc and free tho.

[math]prove((A,B),UnExp,Lits,FreeV,VLim) :- !,
prove(A,[B|UnExp],Lits,FreeV,VLim).
prove((A;B),UnExp,Lits,FreeV,VLim) :- !,
prove(A,UnExp,Lits,FreeV,VLim),
prove(B,UnExp,Lits,FreeV,VLim).
prove(all(X,Fml),UnExp,Lits,FreeV,VLim) :- !,
\+ length(FreeV,VLim),
copy_term((X,Fml,FreeV),(X1,Fml1,FreeV)),
append(UnExp,[all(X,Fml)],UnExp1),
prove(Fml1,UnExp1,Lits,[X1|FreeV],VLim).
prove(Lit,_,[L|Lits],_,_) :-
(Lit = -Neg; -Lit = Neg) ->
(unify(Neg,L); prove(Lit,[],Lits,_,_)).
prove(Lit,[Next|UnExp],Lits,FreeV,VLim) :-
prove(Next,UnExp,[Lit|Lits],FreeV,VLim[/math]
testing

>>9436016
>if you know java

Fuck off code monkey.

>>9436059
Rude, I personally dont use it but Java is a legit choice. What you prefer, python?

>>9417423
what the fuck is that bottom row there for

How strong do I need to be in Python/SQL/R to get a job in data science, even just as an analyst?

Why is Haskell so popular if you dont see it use it that much? And other question, what is Haskell good for?

Why is Haskell so popular if you dont see it used that much? And other question, what is Haskell good for?

>>9436076
No comment on the others, but Perl does have PDL and so deserves to be placed wherever Python is.

>>9436752
Agree, I watched a lecture where says PDL is faster than python.

>>9417423

Shit for brains.

Has programming help with your math skills? Or is any program lie MATLAB improve your understanding in any way?

>>9436752
yeah but science and research is supposed to be open, you can not say that about a perl program (even if you have the "source")

Has programming help with your math skills? Or has any program like MATLAB improve your understanding in any way?

>check wiki
>they say to start with C++
>the recommended textbook is 1000+ pages long
>still have to learn other programming languages and Mathematica
Wew

>>9436752
Perl is unreadable. You may as well be coding in APL.

>>9435795
Matlab does not have this problem

>>9437608
>>9437615
Stick with python brainlets.

>>9434173
>implying it's not highly used
let's say that you program your high performance C++ code to run your calculations, and you are left with 100 Mb of data to graph
What do you do, do you spend half a workday using C++ to get your graphs done, or do you run a python script to plot that shit in minutes? Python has its uses

>>9437624
>shilling python

>>9437615
Perl's not that bad
It's a little unintuitive sure, and it's used less and less for just that reason, look at Perl 6
Python has serious problems, but it's still a great scripting language for research
Can't imagine doing all my scripting in something like sh or Perl

>>9437627
says who? you, against the majority of the scientific community? why are you so insecure that you feel threatened whenever somebody brings up an use for something you don't like?

>>9437660
>the majority of the scientific community
The burden of proof in on you.

>>9437660
You know what, you are taking this too personal. Are you a developer?

Is there something to program cellular automatas with? Like conway's game of life https://www.youtube.com/watch?v=xP5-iIeKXE8

>>9437885
https://en.wikipedia.org/wiki/Core_War

>>9438196
looks neat but what is it?

>>9438206
A sample-efficient global optimization algorithm.

>>9417423
>go on /sci/ programming threads
>everyone says learning c++ first is a great idea
why? what are these new programmers implementing here that needs such fine performance?

>>9438241
with C you actually learn to program. that's the point.

>>9438196
This makes me think of a tree.

>>9435795
Use these arrows.

>>9417482
Haskell

>>9438241
It's a language that shows what's happening like C but has nice higher level stuff when you don't need it.

>>9437624
No, you use Matlab to plot it. Math in python is an ugly hack.

>>9434151
>I don't know what the right tool for the job means: the post

>>9434151
>implement everything by hand in python
>refactor it to numpy afterwards

>>9434107
I've used it to write a simple program with networking to experiment with their gui interface.

The language is neat, you can find how to do everything with the standard library and can prototype fast but lisps don't seem to be very natural at representing things like formulas due to reverse polish notation. It can probably be done with some well designed macros but why even bother, python is way more popular, nicer syntax and easy to write modules in C

whats your latex formatting settings for a typical essay or science paper?

i use \documentclass{article} but the default margins are too big

i know about the geometry package, what are good settings for it?

>>9439933
I personally set my margins to 1.0in seems pretty standard to me. I have messed with having the top and bottom at 0.5in an the sides at 0.8in though.

I'd say it's quite subjective desu - as long as it's readable

What is the most fringe special-snowflake scientific computing language I can learn that will help differentiate me from the pleb masses?

Bump limit reached, time to make a new thread.

NEW THREAD
>>9441598
NEW THREAD
>>9441598
NEW THREAD
>>9441598
NEW THREAD
>>9441598