An implementation of the R Language in Java | Hacker News Reader

An implementation of the R Language in Java

github.com

42 pointsghosthamlet14 comments

Ye Gods! Do you mean to say that this is not an implementation of x in Javascript? [1]

[1] Values for x might include Linux, the XBox, the Apollo Guidance Computer, Intercal, Apache, the French language, the Haskell compiler, Hacker News, Call of Duty, iOS, and Javascript.

https://code.google.com/p/renjin/

From a quick glance I see the following differences:

Renjin

  - development started in 2011
  - GPL v3 licensed
  - working on a custom JIT compiler

fastr

  - development started Aug 2012
  - Apache v2 License
  - uses Truffle and the Graal JIT

Indeed. Would be nice to see how they compare in completeness and performance.

> "FastR is can run the R implementation of the Language Shootout Benchmarks."

Can someone show us the results of these benchmarks? As recently discussed in https://news.ycombinator.com/item?id=5565056, http://benchmarksgame.alioth.debian.org/ lists only the most popular implementation of each language, so it seems like they'll never show up there.

So do either fastr or renjin allow the use of R packages that use C or Fortran code? So much of the value of R is locked in packages. It is kind of the same issue as PyPy, only more so with R.

why not use something like incanter?

My exact thought. R has great tools but the language is wonky. That's not a dig against them. It wasn't designed by language designers. However, its interactive nature is great for data scientists. So it seems like Clojure (possibly Scala) is a better fit than Java.

it isn't the R syntax, as a language, it is rather the package ecosystem and the invested resources it encapsulates.

The ecosystem didnt come into existence by magic. The early adopters did have a choice into which language to invest their time. Obviously the R syntax alone was enough of an incentive to starts building useful libraries in it, instead of lisp, python, etc.

I'm confused. Are you saying that the package ecosystem is what makes R attractive, or that it's what makes it inscrutable?

I'm sure he's saying that package ecosystem is what makes R attractive.

I'd need to dig into the source code of this project more to be convinced, but I trust the implementation and algorithm choice in core R way more than any other stats package except possibly SAS: among other reasons it has far more of the right eyeballs looking at the code. Picking numerically reliable implementations of most estimators is not trivial without the right training. Now, I don't know how much needed to be rewritten for this project, so that's just my general opinion.

Its easy to get addicted to R syntax / capabilities. (Read: Data scientists are lazy)

Reply on news.ycombinator.com