[R-SIG-Mac] New user question

stefano iacus stefano.iacus at unimi.it
Wed Mar 28 22:26:51 CEST 2007


Hi All,
R.app, as Jorg says, has very-very limited apple script support. I  
originally implemented it and it is very weak (for several reasons,  
e.g. no feedback is returned from the executed command; no synch, etc)
Any contribution to add more scriptability to R.app is welcome.

If it is just a matter of running R scripts, then going directly to  
R.app :File-> Source file is enough. You can also configure R to  
source scripts by default on drag&drop, which should be also an  
interesting shortcut.
Of course, you can create an applescript to source a script into R,  
i.e with something like "tell R.app cmd source(file)"

The tcltk package might be a solution but you need the user to  
install X11, which is usually not installed by default of OS X machines.

An alternative is to create a Cocoa app on top of the cmd line  
version of R for each of your ad-hoc scripts. There are several  
examples of Cocoa GUIs built on top of unix commands (i.e. mplayer,  
ffmpegx, etc)

stefano


On 28/mar/07, at 20:19, Jörg Beyer wrote:

> Hi Linda,
>
> I think there is currently no direct, easy and/or satisfying  
> solution for
> your scenario. The probably most convenient solution is to set up your
> scripts in a way that they can be "source"d in via the menu command  
> "File >
> Source File...", which gives you a file selection dialog; R.app  
> will then
> generate a source()-command, read the selected file and execute the  
> found
> code. This is the same as _manually_ typing "source("path/to/R- 
> script")" in
> R.app's console window.
>
> As for Automator, AppleScript and friends, the big picture is a bit  
> more
> complex (apologies if you already know that): The question is not  
> in the
> first place, whether Automator or AppleScript can accomplish this  
> or that,
> it rather depends on the scripting support an application offers.  
> It's a
> design decision made by the application's developer(s). No  
> scripting support
> in the application -- no scriptability; weak scripting support -- weak
> scriptability; ...
>
> So the question is "Does R.app offer scripting support, and to which
> extend?"  A look at R.app's AppleScript dictionary shows that the only
> command that could be useful here is "cmd" to send and execute R- 
> code. There
> is no "source file" command or something comparable in R.app's  
> scriptability
> features.
>
> You could of course develop an AppleScript-, AppleScript Studio- or
> Automator-based solution that simply uses R.app's "cmd" to build  
> and send
> "source("path/to/R-script")" commands and wrap some GUI elements  
> around it.
> But I doubt that such a solution would be worth the effort, and of  
> course it
> depends on the skills you already have with one of Apple's automation
> technologies.
>
> BTW, some month ago I suggested to add a configurable script  
> palette to
> R.app's feature set (similar to the palettes in TextWrangler and  
> BBEdit),
> which would provide a way to easily select and execute script  
> files. We will
> have to wait and see what the future brings. It depends on whether  
> or not
> the developers of R.app find the idea useful, and of course whether  
> they
> find the time to implement it.
>
> Hope this helps.
>
> Jörg
>
>
>> Dear R Mac developers group,
>>
>> I hope this question is appropriate for this list.
>>
>> I'm a new user (a few months) and R has been great. Thank you!
>>
>> If it's possible, I'd like to make a simple GUI or other front end  
>> to a
>> few of my R scripts for my non-coding colleagues to use (or even  
>> little
>> stand-alone apps.)
>>
>> Could Automator or AppleScript, or something else, be used to do  
>> this?
>> (I don't know either yet; looking at Automator it wasn't immediately
>> clear to me whether or not it can talk to R.) Any recommendations
>> appreciated!
>>
>> Many thanks,
>> Linda Palmer
>
> ============================================================
>
> Jörg Beyer
> e-mail:  Beyerj at students.uni-marburg.de
>
> PHILIPPS-University Marburg
> Dept. of Psychology
> Germany
>
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac



More information about the R-SIG-Mac mailing list