[R] R in batch mode packages loading question

Prof Brian Ripley ripley at stats.ox.ac.uk
Mon Nov 7 17:08:17 CET 2011

On Mon, 7 Nov 2011, PALMIER Patrick (Responsable de groupe) - CETE NP/TM/ST wrote:

> Thank you for your response, but this is not exactly what I need
> We are working on a tool that automatically generate R scripts adapted
> to our surveys databases.
> When we want to do a table, we select interactively fields, associated
> labels, functions for an automatic crosstable for example
> Then, the tool executes R in batch mode with the script produced by the tool
> The "many scripts" correspond to the many crosstables people want to do.
> It is not possible produce these different scripts and execute all of
> them in a unique script with source.
> For example, people want to do a crosstable, the tool produce a r script
> that generate the result. Then, the user want to do the same crosstable
> but on a subset or with different labels. Then, the tools produces
> another r script, and execute R which reloads at each time all the
> neeeded packages.
> Is there a way to avoid this thing and to load packages only once?

You have to load them once per R session.  So you could organize your 
work to use a single session, as Duncan suggested.  And that session 
could wait for input (using something like Rserve).

If your packages are taking too long to load, fix the packages. 
Well-written packages load in milliseconds: after all R itself manages 
to load several large packages in 100ms or so.

> Than you in advance
> *Patrick PALMIER**
> ***
> Le 07/11/2011 15:20, > Duncan Murdoch (par Internet) a écrit :
>> On 07/11/2011 5:49 AM, PALMIER Patrick (Responsable de groupe) - CETE 
>> NP/TM/ST wrote:
>>> Hello,
>>> I use R in batch mode. Each time, I execute a script, R is loading each
>>> packages I need in my script. That's Ok
>>> But, I had to execute many scripts , and each time R is re-loading the
>>> corresponding packages, which take to much time
>>> Is it possible ask R to load the packages only once, and stay in memory
>>> in background for further scripts, which would avoid to load the
>>> packages in each script, or if you have another solution that need to
>>> only load packages once in the first scripts, so that further scripts do
>>> not need to load these packages too.
>> Write one script that has a sequence of calls to source() to run the other 
>> scripts.
>> You'll need to be careful that unintentional leftover objects and settings 
>> from one script don't affect the others; you may also want to use the 
>> "echo=TRUE" option when you source, so you see the commands as they are 
>> executed.
>> Duncan Murdoch
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

More information about the R-help mailing list