Is this worth learning? How does one go about learning it and how long would it take?
>>8098767
Don't learn too much of it, but yes, it's a great tool.
It's a shitty language, but it has lots of packages and makes graphing pretty easy once you get used to the awful language conventions.
You should maybe spend <10hours learning basics. Maybe even less. Then pick up the rest as you go.
>>8098767
Horrid language but very large and useful code base.
Worth knowing and using, not really worth creating new things in.
Note that some of the great R stuff there (like most of the nonlinear regression packages) is actually just a wrapping for FORTRAN implementations.
>>8098790
what other languages are better, spss?
I happen to really like the language. It has a lot of convenience for manipulating and exploring data, and the data frame format is actually pretty nice.
I see it as kind of what I've always wished Python would be: high-level, lots of convenient abstractions and syntactic sugar, but without implicitly meaningful whitespace and a community hellbent on being Pythonic.
Don't use RStudio, though. The lag on the REPL response to input, even on simple evaluations, is horrendous and the built-in editor doesn't offer any features you can't find elsewhere with more.
>>8098780
>It's a shitty language
>>8098790
>Horrid language
t. retarded engineers
>>8098825
I bet these faggots like Matlab or Python.
Yes, It's worth learning and it won't take long to start getting work done, but actually getting a good understanding of how it works will.
I would also recommend that you learn how to use the plyr and dplyr libraries as soon as possible.
>>8098847
No, R really is a horrid language.
http://r.cs.purdue.edu/pub/ecoop12.pdf
>>8098856
No, it's not. That paper doesn't support your statement either.
>>8098869
>That paper doesn't support your statement either.
It blasts it left and right on everything from terrible performance to incoherent design and "massively inefficient" implementation.
>"As a language, R is like French; it has an elegant core, but every rule
comes with a set of ad-hoc exceptions that directly contradict it."
>>8098888
>It blasts it left and right on everything from terrible performance to incoherent design and "massively inefficient" implementation.
You're confusing language design and implementation.
>>"As a language, R is like French; it has an elegant core, but every rule
>comes with a set of ad-hoc exceptions that directly contradict it."
That doesn't mean it's bad.
Also, checked.
>>8098898
>You're confusing language design and implementation.
So did the creators of R since there is no specification other than the implementation itself.
>>8098898
>That doesn't mean it's bad.
No, the lack of support for multi-threading, two inconsistent (and both shitty) object systems, and horrible performance mean its bad.
The surface things like the syntax and basic types I actually like.
>>8098846
Should i use 64 bit R then?
It's real handy for personal use, but I imagine it's a hassle for more computationally demanding bigger professional data.
>>8098918
I usually use Rgui for prototyping and run things as complete scripts from the command line.
>>8098910
Is it not the best biostatistical tool out there though?
>>8098944
Best does not imply good.
It's best because of the code base, not the quality of the language.
>>8098910
Most of the paper's objections boil down to "you'd be stupid to use it as an enterprise language" when that was never, and still isn't, the point of the language, and almost all of their evaluations of laziness, dynamic programming, etc., are an evaluation of the programmering community, not the language. It's not surprising that programmers in R don't take full advantage of the complex features of the language, since they tend to be scientists with at-best amateur programming skill, not professional programmers.
R is meant to be used for quick research, not large applications. It excels at what it's meant for, and of course it's flawed when you try to fit it into applications it wasn't developed for.
>>8099041
>"you'd be stupid to use it as an enterprise language"
Why are you using quotation marks around your loose misrepresentation of the paper?
> almost all of their evaluations of laziness, dynamic programming, etc., are an evaluation of the programmering community, not the language
Outright lie. The community is only discussed in the conclusion, and it is there in the context of explaining the reasons for the proliferation of R despite its flaws.
R is gimped by a very inefficient implementation and lack of crucial features like multi-threading. Just look at the performance benchmarks in the paper, where its performance lags behind Python by an enormous degree (43 times slower in standard benchmarks).
>R is clearly slow and memory inefficient. Much more so than other dynamic languages. This is largely due to the combination of language features (call-by-value, extreme dynamism, lazy evaluation) and the lack of efficient built-in types. We believe that with some effort it should be possible to improve both time and space usage, but this would likely require a full rewrite of the implementation.
The only reason for using R is inertia due to the usefulness of the existing code base. That is its only added value over alternatives like Python.
>>8099130
>The community is only discussed in the conclusion
Wrong. All of the non-Shootout comparisons and analyses are built on tests of their cumulative codebase, which is taken from user-submitted packages on CRAN/Bioconductor vignettes and user-submitted code. See: Figures 6, 8, 10, 11, 12, 13, 14, 15, and 16.
>Just look at the performance benchmarks in the paper, where its performance lags behind Python by an enormous degree (43 times slower in standard benchmarks).
And both of them are enormously slower than the same benchmarks established in C. That doesn't mean we should all move to C. Not to mention that for most users of the language, the difference in run-time between Python and C is on the order of several milliseconds.
You're absolutely correct, R is less efficient than other languages and lacks some features. What I'm saying is, you're the only person who cares.
>>8098910
You're absolutely correct, but you're probably wasting your time trying to explain it to this guy. People who try to defend R as being a good language generally don't have the background to understand why it is such a shitty language.
