R or Python for data science Veeky Forums?

Fortran

>Unfortunately I don't have a CS degree
>Unfortunately

It's a plus.

how do you figure? I wish I had a CS degree so I could actually get a fucking job

both, and maybe a few others

shhh it's a sci circlejerk.

>I don't have a CS degree
>I used it during undergrad for research
????????

I used data analytics like R, python, and a bunch of stats. Im wondering if I need a CS degree to validate my skills for employers

It's just a mathfag scared as fuck of not getting a job after graduating.

same dawg

Wolfram Mathematica.

Python

because you can also do actual work with it

you're ugly

everyone should know python

end this fucking meme. python is a language for normies and brainlets who cant learn C/C++.

Attached: CS degree jobs.png (823x378, 56K)

Why the fuck would I use C for most data processing applications? Fucking embedded systems developers, the people who use C day in and day out, wouldn't even do that. Why not just write it in x86 assembly while you're at it?

90% of people on Veeky Forums don't need whatever minor speed gains you get from C vs well-written python, because there's no difference between code that runs in 3 ms and code that runs in 5.

This
R is a mess to get into at the beginning though, you'll be using help() a lot, but once you get the grip it's pretty useful. Also, it works great for quick calculations that are too simple to code in depth but too complex to simply put them on a calculator.
Wouldn't that apply to pretty much everything?

X

Haskell

>Wouldn't that apply to pretty much everything?
Yes.

Attached: 1520773733343.jpg (982x1378, 166K)

Julia

The main reason i still use R despite using mostly python is ggplot2.

Use Matlab, much better graphs.

/g/ and ML PhD here.
Most opinions posted in the thread so far are shit and you should learn Python 3.

>R
Scales like ass

>C++
Let skilled programmers deal with it, you can rely on their work through most python libraries (numpy, tensorflow, pandas, scikit, etc)

>MATLAB
Walled garden for faggots

>Anything else
You're being troled XDXD

>but muh performance
See comment on C++

>muh CS is /g/ay meme
I'll be wiping my tears with all the money I make

Attached: 9fxvy22dypsy.jpg (1334x750, 51K)

>R is a mess to get into at the beginning though
what the fuck? This is what happens when you don't weed out brainlets by teaching programming first with C++ or Fortran (yes, in my Math degree we did it)
R and python are easy, and using the documentation is the INTENDED way of using a language you retard, good luck trying to properly do a templating class in C++ tho.

mmm in my company we use R to deploy cloud based analytics...

Google uses Python in most projects and data driven sites like YouTube and Plebbit use Python. It's not that you can't use R for big projects, it's just that it's a shitstorm to do it properly (also because you'd have to integrate it with whatever language you use for front end)

I see. I am a Junior still so idk many things.
I guess i will just grind 3 years exp then switch to another place

I'm trying to become a data analyst/scientist, would a master's degree in applied statistics be useful in this regard? I currently have experience in R and Python in research/ analytical applications from my undergrad.

Attached: fullsizeoutput_2b10.jpg (998x802, 124K)

Found the biggest cuck on this board

Yes, or possibly something fully data science oriented (there are a few masters around now)

No point in writing 50k lines of code that have already been written by teams of professional C++ devs unless your time is worth nothing, you fucking nigger.

I said at the beginning, that means with no knowledge of any command of R. Any new programming environment has a certain degree of difficulty even if you know how to program simply by the fact that you have to get used to it. Of course once you get the idea of how it works, it's pretty simple and useful, which I mentioned. You didn't read the rest of my post user.

>>muh CS is /g/ay meme
>I'll be wiping my tears with all the money I make

Making money has nothing to do with whether cs is trash or not. Business majors and MBAs make money too while learning nothing.

I agree. It's easy to do some simple beginner stuff like doing hypothesis tests or whatever, but medium level stuff like implementing some algorithm is just ass.

R

Why do people say R has bad syntax or is ugly? It's a high level language with incredibly similar syntax to Python. The biggest difference is R doesn't have the list comprehensions of Python, everything else is just cosmetic differences. Hell, half the time people recommend R they rec it in combo with numpy/pandas, which has damn similar data structures to R's data frames.

I honestly get the feeling most people who say R is ugly just can't get past using

which you also don't even need to do, R supports = just fine. The only real reason to use

There are times when it's important to have speed. And that's when you use 4tran,

Has nothing to do with that, it's just that the syntax is way more intuitive in python. It's probably better for e.g. hypothesis testing but as a general data analysis language python is definitely better.

This, fuck python.

>There are times when it's important to have speed.
If you're capable of writing actually efficient code in even a single language then you're not in a position where you need to ask "R or Python for data science Veeky Forums?"

>teams of professional C++ devs
t. pajeet

Excel is for small datasets. R can handle large data frames no problem, but depending on what you're doing, it can be very slow. I don't know Python is much faster.

What about Haskell? It totally makes sense for data analysis. Something like it will probably be the future.

Coding is not hard
Writing good code IS hard. The main objectives are writing readable and performant code, and many people who actually were trained at this still struggle to do so - how well do you think someone untrained would do?

The real answer

Neither. Use Matlab.

Gnuplot and perl if you like science

>PyTorch
>Tensorflow
>Scikit
>Keras
>Theano

Meanwhile, in R:
>absolutely nothing

Clearly Python OP.

Attached: 1502292568934.jpg (1008x343, 42K)

Python

They are both bugman languages. Don't be a bugman.

R has interfaces to tensorflow and keras. No single package replicates scikit or theano but there are many packages for various types of machine learning.

Gnuplot sucks tho

Pic semi related

Attached: 1509416654112.png (429x369, 161K)

Fucking kill yourself
derpibooru.org/1257982

Attached: Bronies, this is your mindset.jpg (307x156, 24K)

But you are the greatest cucklord of all time by using R.

No, man up and do everything in assembly/byte code.

Also, Perl is jank.

>What about Haskell?
wiki.haskell.org/GHC/Memory_Management
>it's not uncommon to produce 1gb of data per second

The research computing department at our uni did a survey, Python is the clear leader, however R is a strong second place. R is pretty much ubiquitous in genetics, biology and pharma sciences here. Both R and Python have a ginormous ecosystem built around it, and most times the speed of implementation/prototyping is more important than execution speed. Fortran is very widely used still, especially in CFD (and related subjects) where raw performance is key.

If you could actually code it wouldn't matter what programming language you do it in. Most of people asking these questions are about to start learning, but will never follow through with it.

The programming language you choose comes down to either preference, or what your peers are using.

Attached: irritation.jpg (250x250, 7K)

>deep learning
cutting edge stuff user. bet you did a lot of learning with that python library

R for statistics, Python for deep learning.
I prefer R, but it's a matter of taste

I literally just came to sci to find another faggy thread about data science. This field is dead. Period. There are already way more potential "data scientists" that required. Especially if you are a white male your chances to get a job are almost 0.

R has more stuff that python. You clearly never tried to do something serious.

Attached: 1518346731439.png (644x800, 15K)

>gay anime yaoi
Sums /g/ pretty well.

Attached: no way.jpg (641x530, 25K)

>Serious work
>Learning one programming language

Listen OP, you're like a guy who wants to be a carpenter asking if he should learn the hammer or the saw.

>these people call themselves CS majors
>he unironically doesn't understand the importance of linguistics in CS
are they just handing out CS degrees now?
Are tards this dumb to not know that SWE is a subset of CS?

>not having 32GB of ram anyways

Attached: sci in a nutshell.jpg (646x974, 84K)

What is that picture supposed to show? Lying with statistics? Amplitudes in QM? Gender wage gap? All three?

Hah how's it feel to be a RAMlet. My desktop has 256GB and 56 cores because the business office gave me a retardedly high budget for it. Granted I actually use all of it in fluid structure interaction simulations.

Yes.

Attached: j4pwao8j3.jpg (610x780, 94K)