8 ms·

Julia looks amazing BUT as an R user: 1) RStudio is just such an amazing tool with rmarkdown and pandoc. I run my statistics in a document and can output to ht

by baldfat 12y ago

Julia looks amazing BUT as an R user: 1) RStudio is just such an amazing tool with rmarkdown and pandoc. I run my statistics in a document and can output to html, pdf or Word. I can't tell you how important the ability to print out to Word is in my corporate setting. 2) The libraries and tools are vast (Seems like Linux tons of options but really one or two are perfect for the individual use) 3) 99% of what I do takes less then 2 seconds with R Personally if I have a need for large datasets I will look to Julia if I can't get a solution with R and data.tables.

johnmyleswhite 12y ago
Personally, I don't see Julia replacing R for a while. As you note, the R community has much of what it needs already. When I initially switched from R to Julia, it was because R wasn't really usable for the kind of work I do. It's since become clear to me that I work on problems that don't come up for most R users: problems involving large sparse matrices, including large-scale optimization and MCMC. If you don't deal with those kinds of things, R is a good choice.
- bachmeier 12y ago
  I think the best route is to create shared libraries using Julia and call them from R.
- peatmoss 12y ago
  It seems like MCMC is a potentially killer app for Julia. Any thoughts as to how Julia's MCMC facilities compare to, say, something like Stan?
  - johnmyleswhite 12y ago
    I think MCMC.jl is pretty immature compared with Stan. But there is work on a Stan wrapper. Personally, I mostly write MCMC code by hand these days.
    - peatmoss 12y ago
      Granted I don't have much experience with Stan, but from what little I've poked around RStan, the workflow left something to be desired. Setting up a C++ environment, and then embedding Stan code (which appears to be a C++ DSL of sorts) as strings inside your R code seems... unpleasant. Perhaps if MCMC.jl matures, and if it can offer performance competitive with BUGS, JAGS, Stan, etc., I could see Julia's statistical fortunes rise along with Bayesian methods generally. I'm working on a PhD in a discipline that is just now beginning to dip its toes in Bayesian waters. I get the feeling that the adoption surge is yet to come. Interesting that you're implementing your own MCMC methods mainly. As part of my coursework I did a little bit of that, but perpetually felt like I wasn't smart enough to anything sophisticated--either in terms of exotic sampling methods, or working with very complex posterior distributions. It may just be my frequentist R toolbox experience messing with me, but I know I feel more comfortable with the safety blanket of a framework.
- ViralBShah 12y ago
  Second that. The purpose of Julia is not to replace every other language that people use for technical computing. I personally use Python for other projects (www.circuitscape.org), and only recommend using Julia for new classes of problems that need Julia's performance and capabilities. That class of problems keeps expanding over time.
bachmeier 12y ago
I agree. I don't see Julia getting much traction in the statistics community. There is the legacy thing (the tools, existing code, and knowledge) but there is also the fact that R is designed specifically for statistics. Julia is a good replacement for Matlab, but there is a reason Matlab isn't used much in statistics.
peatmoss 12y ago
I agree on RStudio. I used ESS for a while, but couldn't in good conscience justify it for new users. So, RStudio became my mantra for new R users, and I've followed suit. I've wondered what it would take to instrument support for Julia in RStudio. RStudio is open source, but I don't really know how generically it could interface to whatever repl.
- ViralBShah 12y ago
  RStudio looks really impressive. I have heard reliably that they do use the AGPL license and legal threat to get people to pay up for RStudio licenses. I have nothing against people charging for good software, but this sounds like some sort of bait and switch. I wonder if someone else can confirm if this is actually true.
  - peatmoss 12y ago
    I'm not sure I understand how the AGPL allows them to do that. Doesn't the AGPL just require people "releasing" derived works as part of a service to also make their modifications free? I don't really know much about the AGPL other than what I could find on Wikipedia. Assuming I understand the added requirement of the AGPL, wouldn't RStudio's legal recourse apply only to derived works that didn't release their code?
    - ViralBShah 12y ago
      I wonder if any code written in RStudio gets contaminated with AGPL. I also believe that many companies will just pay the fees out of fear of litigation.
      - mweylandt 12y ago
        Why would you think anything of the sort? Is everything written in Emacs GPL? Is everything written in Eclipse under the Eclipse license? Like (almost?) all FLOSS dev tools, there's no connection between how the code is written and the ensuing license. Suggesting such (even by way of a comment) is rather unfair.
idunning 12y ago
While it is a nice tool, RStudio is AGPL, which means it can't be used in many companies due to contamination fears, including Google but I understand many other companies as well.
- baldfat 12y ago
  Well I would think they would pay for RStudio server. I know $10,000 seems steep but I am sure Google could afford that.

by baldfat 12y ago

johnmyleswhite 12y ago

Personally, I don't see Julia replacing R for a while. As you note, the R community has much of what it needs already. When I initially switched from R to Julia, it was because R wasn't really usable for the kind of work I do. It's since become clear to me that I work on problems that don't come up for most R users: problems involving large sparse matrices, including large-scale optimization and MCMC. If you don't deal with those kinds of things, R is a good choice.

bachmeier 12y ago
I think the best route is to create shared libraries using Julia and call them from R.
peatmoss 12y ago
It seems like MCMC is a potentially killer app for Julia. Any thoughts as to how Julia's MCMC facilities compare to, say, something like Stan?
- johnmyleswhite 12y ago
  I think MCMC.jl is pretty immature compared with Stan. But there is work on a Stan wrapper. Personally, I mostly write MCMC code by hand these days.
  - peatmoss 12y ago
    Granted I don't have much experience with Stan, but from what little I've poked around RStan, the workflow left something to be desired. Setting up a C++ environment, and then embedding Stan code (which appears to be a C++ DSL of sorts) as strings inside your R code seems... unpleasant. Perhaps if MCMC.jl matures, and if it can offer performance competitive with BUGS, JAGS, Stan, etc., I could see Julia's statistical fortunes rise along with Bayesian methods generally. I'm working on a PhD in a discipline that is just now beginning to dip its toes in Bayesian waters. I get the feeling that the adoption surge is yet to come. Interesting that you're implementing your own MCMC methods mainly. As part of my coursework I did a little bit of that, but perpetually felt like I wasn't smart enough to anything sophisticated--either in terms of exotic sampling methods, or working with very complex posterior distributions. It may just be my frequentist R toolbox experience messing with me, but I know I feel more comfortable with the safety blanket of a framework.
ViralBShah 12y ago
Second that. The purpose of Julia is not to replace every other language that people use for technical computing. I personally use Python for other projects (www.circuitscape.org), and only recommend using Julia for new classes of problems that need Julia's performance and capabilities. That class of problems keeps expanding over time.

bachmeier 12y ago

I agree. I don't see Julia getting much traction in the statistics community. There is the legacy thing (the tools, existing code, and knowledge) but there is also the fact that R is designed specifically for statistics. Julia is a good replacement for Matlab, but there is a reason Matlab isn't used much in statistics.

peatmoss 12y ago

I agree on RStudio. I used ESS for a while, but couldn't in good conscience justify it for new users. So, RStudio became my mantra for new R users, and I've followed suit. I've wondered what it would take to instrument support for Julia in RStudio. RStudio is open source, but I don't really know how generically it could interface to whatever repl.

ViralBShah 12y ago
RStudio looks really impressive. I have heard reliably that they do use the AGPL license and legal threat to get people to pay up for RStudio licenses. I have nothing against people charging for good software, but this sounds like some sort of bait and switch. I wonder if someone else can confirm if this is actually true.
- peatmoss 12y ago
  I'm not sure I understand how the AGPL allows them to do that. Doesn't the AGPL just require people "releasing" derived works as part of a service to also make their modifications free? I don't really know much about the AGPL other than what I could find on Wikipedia. Assuming I understand the added requirement of the AGPL, wouldn't RStudio's legal recourse apply only to derived works that didn't release their code?
  - ViralBShah 12y ago
    I wonder if any code written in RStudio gets contaminated with AGPL. I also believe that many companies will just pay the fees out of fear of litigation.
    - mweylandt 12y ago
      Why would you think anything of the sort? Is everything written in Emacs GPL? Is everything written in Eclipse under the Eclipse license? Like (almost?) all FLOSS dev tools, there's no connection between how the code is written and the ensuing license. Suggesting such (even by way of a comment) is rather unfair.

idunning 12y ago

While it is a nice tool, RStudio is AGPL, which means it can't be used in many companies due to contamination fears, including Google but I understand many other companies as well.

baldfat 12y ago
Well I would think they would pay for RStudio server. I know $10,000 seems steep but I am sure Google could afford that.