>He doesn't program in R >He thinks he can produce

>R
>programming

>>60657996
ok, smartass.
plot me a histogram with specified number of classes, without using additional packages

>>60657996
(You) 4 d effort

>>60657996
>He probably thinks that R isn't a real programming language
R is a real programming language in the sense in that JS and PHP are real programming languages.

I'm not a statistician, and creating Excel charts from data that already has been aggregated by the mainframes is enough to satisfy my boss's visualization needs.

>>60658396
JS and Python have great dataviz libs tho

>>60658560
LOL Python's Seaborn is a clone of ggplot2

File: 1496005795031.jpg (147KB, 1000x1000px) Image search: [Google]

147KB, 1000x1000px

>>60657996
>he wasted time learning R instead of a more general programming language
>he didn't realize that all the useful libs in R have ports or equivalents
>he now has to make a thread on /g/ to try and salvage his ego

>>60657996
mercy best grill

R is the worst language I've ever used.

>>60657996
I learned R a few months because I want to get into data science. It's a great language for ML/DS/AI, especially with all the packages available on CRAN, but as a general programming language, it's weak.

>>60657996
While I use R, I wouldn't call it a programming language, it is more like a statistics tool.
You can make functions, but you wouldn't make larger programs in R.
Every time I have done that, I immediately notice the 3x increase in runtime and extra use of memory.
If something as essential as for loops is slow and causes memory leaks, I wouldn't make programs with it.
It is great for plotting data and making analysis.

R isn't a language it's a letter.

Dumbass.

I don't know R, but I can give an example of real world use. I was on a contract at a major pharmaceutical company, the one with the flower name, and they had a massive parallel computing cluster built on AWS that was doing protein simulations for drug research. R was heavily involved, although I'm not sure if R was actually running the simulations or was being used to analyze the results. I'm guessing and i hope the latter.

File: yoyo_think_anim_small.gif (3KB, 50x50px) Image search: [Google]

3KB, 50x50px

Is R the same thing ad Rust?

>>60660336
>I learned R a few months because I want to get into data science
Then you fucked up, because no one in industry actually uses R. It's Python everywhere.

File: 3.png (15KB, 112x112px) Image search: [Google]

15KB, 112x112px

>learned specialized language instead of a more general one
Hurr durr

>>60658560
>dataviz

>>60660247
Pleb tastes.

>>60657996
why should i use r over python?

File: image.png (167KB, 396x385px) Image search: [Google]

167KB, 396x385px

>>60665328
Data visualization

File: argument_pyramid_of_shame.png (71KB, 707x530px) Image search: [Google]

71KB, 707x530px

>>60665414
thank you for the smart well toughtout answer

File: 1474576536420.png (452KB, 785x757px) Image search: [Google]

452KB, 785x757px

>>60665445

Many of the reasons to use R are many of the reasons many people bash R

Non standard evaluation and ambiguous syntax:
R is more pythonic than Python!
If you use pandas you have to use so much syntactical fluff to clarify your intent that it makes a clear train of through look convoluted.
the ability to input fields without quoting them and having R figure out if it's a field or a variable makes the syntax much terser and easier to read

every package in the global namespace
R is built for ad hoc analyses and in such analyses I don't want to have to prefix pd, np or plt before every function or import it explicitly first

Other syntactical choices like the tendency to have very short function and variable names make writing and reading code faster. The domain is typically well enough defined to make the intention clear

custom operators, most prominently the magrittr pipe %>% allow you to chain everything. A functional approach with pipes makes data analysis much easier to read and you don't need special classes with their own methods for it.

these are small gains but they make the workflow so much more intuitive

on plotting:
imo ggplot2 has the best balance between usability and feature richness, the grammar of graphics way of doing things is genius, Python alternatives are either more verbose or ugly and anyone who claims the ggplot2 python port is "basically the same" has never used ggplot2 properly.

There are more but it's late and I'm tired.
Source: I'm a data scientist who uses R every day.
Before you start bitching: I know Python, I use Python, it's great, it's just not my first choice for data analysis

>>60665630
sounds like you are just too lazy or stupid to learn to use python to be honest

File: 1388412490321.gif (732KB, 360x203px) Image search: [Google]

732KB, 360x203px

>>60665630
ah ok. sounds reasonable.

>>60665720
Did you even read my post? I use python

>>60657996
>using R
S as in Senpai

>>60658017
yes

>>60658021
>without using additional packages
I don't even use R, but what's up with this retarded arbitrary restriction?