/sci/ - Science & Math

[Anonymous No.10721836]

Redpill me on R. How much is it used in science today and/or other economics fields or should i just learn python

[Anonymous No.10721838]

>>10721836
Use whatever language you want to. "Which language should I use" is the biggest meme people who never end up programming obsess over.

[Anonymous No.10721850]

Learn both. Takes like 30 minutes each, ffs.

[Anonymous No.10721855]

>>10721850
>30 mins
are you kidding me

[Anonymous No.10721877]

>>10721855
>are you kidding me
He probably is.

Do you need R or Python specific packages? Which software does your field mostly use? (For instance, statisticians tend to work with R, ML fanboy tend to work with python, Control engineers work with matlab, people in stochastic programming (supposedly) work with Julia).
I believe these are the most general questions you may want to answer to make your choice

If you can't answer these questions, just go with Python. If you master Python, you shouldn't have much problems with R

[Anonymous No.10721915]

>>10721836
Used it recently in a bioinformatics project, it generally has useful packages that might not be found in python and is great for statistics. Python is good in general and easier to learn. Unless your field requires or benefits from R specifically, go with Python as >>10721877 says.

[Anonymous No.10721925]

>>10721855
Probably meant if you already know how to program lol

[Anonymous No.10721931]

>>10721836
It is essential for quantitative biology especially genetics and ecology

[Anonymous No.10721960]

>>10721836
Dont worry about learning a specific language, just learn a few. You dont even really need to know any of those dynamic languages to write them anyway desu.

[Anonymous No.10722119]

>>10721836
Depending on what kind of job you want to do, Python might be seen as better by your employers so I would advise to learn Python in order to get hired. Afterwards I suggest that you should try to learn a little bit of R. I didn't find it that difficult to learn and it has some packages that can make your life much easier.

[Anonymous No.10722455]

>>10721836
Fortran

[Anonymous No.10722609]

I love being able to use periods in variable names in R. Makes organizing related values so much easier than in any other language that doesnt allow this.

[Anonymous No.10722613]

>>10721836
>Redpill me
GTFO pill-popping /pol/esmoker

[Anonymous No.10722788]

>>10722613
>Le pol
Reddit is on the right

[Anonymous No.10722885]

>>10721836
this is just statistic software comparable to SPSS. I would only use it when testing on data you have created from whatever observation and experiment whether science or economics. The tool is becoming preferable over SPSS this they and age as its fairly open source.

[Anonymous No.10722916]

>>10722788
>Reddit is on the right
yup
>>>/r/the_donald/

[Anonymous No.10722979]

>>10722916
LEWD!!

[Anonymous No.10723087]

>>10721836
Let me redpill you on R
I have worked with R to analyze simulation output for the last year. R has made me an angrier, less patient person. For every SIMPLE FUCKING TASK you want to do in R, there are at least three different notations for doing it, only one of which works, DEPENDING ON CONTEXT

Working with R is like trying to have a discussion with a pedantic autist who also has ADHD and can't stick to one way of doing things at a time. You want to do anything more complex than some vector math? You're going to have to deal with this shit.

Avoid R if you can. Use literally any other language with a stats package. R is not a good language. I will not revisit this software once I have finished this project.

[Anonymous No.10723169]

>>10723087
>For every SIMPLE FUCKING TASK you want to do in R, there are at least three different notations for doing it, only one of which works, DEPENDING ON CONTEXT
Fuck this is dead on... like 100% nailed it.

[Anonymous No.10723209]

>>10721836
i really hate the language but it really cant be beat when it comes down to uncommon stats methods

[Anonymous No.10723332]

R is a dream to use, I love it. It's my favorite language.

>>10722885
only psych major undergrads who refuse to learn how stats actually works use spss

[Anonymous No.10723338]

>>10723087
>allowing the user freedom of expression is a bad thing
spotted the python cultist

[Anonymous No.10723343]

>>10721931
You can use python for everything. R is for fags who can't write their own stuff

[Anonymous No.10723352]

>>10723343
would you cook a meal with a swiss army knife? fuck no. you'd use a dedicated kitchen knife. python is a general purpose computing language but that makes it worse for scientific tasks, not better.

[Anonymous No.10723367]

>>10723087
This is complete bullshit and you know it. If you've actually bothered to read the documentation or a book on R the simple stuff is exactly that: simple. Plus, the degrees of freedom R grants you is only bad if you're a sloppy programmer. Yes, R is sometimes a bit too lax in its interpretation of your code, but I haven't ever had an issue that couldn't be easily resolved within two minutes.

[Anonymous No.10723378]

>>10723367
ok but to be fair the documentation for R can be really shitty. if you're not doing something really simple the examples often don't help, and third party packages are a crapshoot

[Anonymous No.10723428]

>>10721836

The true redpill:

R is valuable, but focus on learning Python. There is no one that I know that principally uses R that also have a good understanding of software engineering practice - y'know, what would actually get you hired. You ought to be able to properly document and modularize your code, make use of version control, write unit tests, etc. While knowing Python does not automatically confer this knowledge, there's a lot more resources for it, and more development teams use Python than R.

[Anonymous No.10723431]

>>10723428
you're absolutely right but most of the people i know who primarily use R are locked into academic tracks and aren't trying to get programming jobs in industry

[Anonymous No.10723557]

>>10721838
trü

[Anonymous No.10724080]

>>10723367
Tell me, how do you access the 4th element of the output produced by a built-in function? Is it
>[]
>[[]]
Or
>$context.dependent.element.name_foundindocumentation
?

[Electric cars or hydrogen-powe(...) No.10724130]

>>10723087
>Let me redpill you
>>10723428
>The true redpill
GTFO /pol/itard fgt pls

[Anonymous No.10724189]

>>10724080
function(arg)[index]

[Anonymous No.10724194]

>>10723378
R documentation is leaps and bounds ahead of others. The quality of the documentation is the best feature

[Anonymous No.10724775]

>>10724080
well, check the docs and see what the function returns. if it returns a vector, []. if it returns a list, [[]]. if it returns a data frame, $

you could also look at the object that's returned first instead of assuming it's one thing and running into errors. program iteratively with prototyping, stop trying to do it all in one pass

[Anonymous No.10724783]

>>10724080
you read the function documentation you fucking brainlet

[Anonymous No.10724797]

>>10721836
I know python, R and MATLAB. I use python the most.

[Anonymous No.10724817]

>>10721850
quality bait.

[Anonymous No.10724819]

>>10722609
you might be the only man glad to have periods. what's wrong with underscores, though?

[Anonymous No.10724865]

>>10721836
I've worked with both for over a decade or so. R has always had the advantage when used correctly and still has some cache in data science circles. However, in more "Pls be patient, I'm a coder/programmer" oriented circles python is totally winning out to the degree that you won't see R at all sometimes. There are all sorts of reasons for these, I would recommend to just learn both. Python is a lot easier now that you don't have the 2 vs 3 bollocks of a few years ago going on so much.

[Anonymous No.10724889]

>>10721850
based. brainlets btfo