News for hackers. Stuff that matters.
Re:Julia is awesome, but.. by tenfingers on Monday December 29, 2014 @10:31AM (#8810140)
I would say Julia's target is much wider than R.

R is actually a nice language (also with multiple dispatch!), but it's basically useless for any actual numerical work. Too slow. There have been efforts in the past to introduce a JIT compiler, but nothing has become mainline yet. The GC is also very poor.

The actual base system in R is also very small. The built-in vector types helped the language mitigate the issue due to the fact that it's very easy to build a C extension that does the actual computation by just taking the underlying memory pointer (it's actually easier to do than python and it maps much better). But it stops there. If it wasn't for the large swath of statistical packages, R wouldn't have much traction. Not to mention that the code quality of the typical CRAN package is quite low, coming from people with a background in statistics and very little CS knowledge.

The hope with Julia is being able to actually develop methods in Julia itself. I personally looking forward to replace python&c++ with Julia&Rust in the future.

Re:Julia is awesome, but.. by bachmeier on Monday December 29, 2014 @11:16AM (#8810339)
> There have been efforts in the past to introduce a JIT compiler, but nothing has become mainline yet

It's been included in the default install of R since version 2.14. You can compile any function you want, and in some cases you can get a speedup of several times. Sure, you're not getting the speed of C, but it's there, it works, and in the future it will lead to much larger speedups.

Some info and timings are here: http://www.rinfinance.com/agenda/2014/talk/LukeTierney.pdf

The experimental version of the compiler can achieve speedups in the neighborhood of 20x.

Re:Julia is awesome, but.. by lottin on Monday December 29, 2014 @02:43PM (#8811419)
Well, it's true, it is included in the default install. After unsuccessfully trying to locate the elusive 'compiler' package at CRAN, I came to the conclusion that the R compiler was an urban legend. Now I realise it was already installed all along from the start and all I had to do is 'library(compiler)'. Silly me.
Re:Julia is awesome, but.. by bachmeier on Monday December 29, 2014 @04:13PM (#8811817)
> Silly me.

Not really. That's an example of one R's rough edges. I did the same thing.

Re:Julia is awesome, but.. by grayclhn on Monday December 29, 2014 @11:23AM (#8810372)
Are you sure you're referring to JIT as opposed to byte-compilation?
Re:Julia is awesome, but.. by bachmeier on Monday December 29, 2014 @02:17PM (#8811282)
It is a byte code compiler that has a JIT mode.[1] I'm pretty sure that's what the parent comment was referring to.

[1] See page 7 of the presentation I linked above.

Re:Julia is awesome, but.. by grayclhn on Monday December 29, 2014 @03:10PM (#8811535)
I didn't know about the JIT mode. Thanks for the specific pointer.
Re:Julia is awesome, but.. by tenfingers on Monday December 29, 2014 @11:31AM (#8810419)
I have R 3.1.x and I never noticed (it's in "compiler"). It's off by default... pretty sad.
Re:Julia is awesome, but.. by jghn on Monday December 29, 2014 @02:10PM (#8811233)
It can, but it often doesn't, at least not a year or so ago (perhaps it got much better). I've seen the examples with great speedups but every time I've tried it on real code I never saw anything. You're much better off just using Rcpp for those sorts of cases.
Re:Julia is awesome, but.. by edraferi on Monday December 29, 2014 @11:37AM (#8810451)
If it wasn't for the large swath of statistical packages, R wouldn't have much traction.

I agree, but Community & Ecosystem are critical to any language's success. R's community is great and getting better [1].

You could make very similar criticisms about JavaScript, which is a lousy language that has accomplished amazing things thanks to third-party extensions and a very active development community.

[1] http://www.tiobe.com/index.php/content/paperinfo/tpci/index....

Re:Julia is awesome, but.. by minthd on Monday December 29, 2014 @11:42AM (#8810474)
Why rust ? can't Julia offer enough speed?
Re:Julia is awesome, but.. by sixbrx on Monday December 29, 2014 @02:19PM (#8811294)
In my experience Julia is quick with low level array oriented programming which doesn't allocate ("C in Julia"), but if the garbage collector is needed within loops or you use anonymous functions there then you'll suffer greatly.
Re:Julia is awesome, but.. by hayd on Monday December 29, 2014 @07:33PM (#8812662)
Have you seen the recent (unmerged) GC work https://github.com/JuliaLang/julia/pull/8699
Re:Julia is awesome, but.. by sixbrx on Tuesday December 30, 2014 @01:13PM (#8815793)
I had not, thanks for the link. I haven't given up on Julia!
Re:Julia is awesome, but.. by tenfingers on Tuesday December 30, 2014 @03:08AM (#8813775)
Rust covers the lower-level much better. In Julia for example you currently don't have a nice way to modify C-like structs. You can offset memory directly, but it's a pointless exercise. Even if a record-like syntax is added, the dynamic nature of Julia might incur in different latencies, you have the GC (not optional), and the runtime is quite big.

Rust is filling the same niche I'm using c/c++ for, where I need absolute control of every detail: not necessarily for performance.