[R] Path to R script

[David Scott]

David Scott wrote:
> Sharpie wrote:
>> Johannes W. Dietrich wrote:
>>> There is certainly a trivial solution for my question, but I can't 
>>> find the answer in the documentation.
>>>
>>> I need a platform independent method to obtain the file path of the 
>>> current R script.
>>>
>>> My working group uses R on several machines with different operating 
>>> systems including Mac OS X, Windows and Linux, and for obvious 
>>> reasons the file hierarchies are very different. Therefore, it would 
>>> be useful to be able to address data files and external procedures 
>>> relative to the path of the script.
>>>
>>> Thank you for any suggestion,
>>>
>>> J. W. D.
>>> -- 
>>>
>> The best recipie I have for dealing with this situation is:
>>
>>   1. Place R scripts in a package (a very good idea anyway if you are
>> distributing code to others).
>>
>>   2. Place a configuration file in the package that contains info such as
>> paths to data directories.
>>
>>   3. Customize the configuration file after installing the package (if
>> necessary).
>>
>> Then your script accesses the configuration file, which is in a known fixed
>> location, and uses the information inside to find other files that may be in
>> unknown and changing locations.
>>
>> You could also place the configuration file in a location such as:
>>
>> ~/.MyPackageConf
>>
>> Since:
>>
>>   path.expand( '~/' )
>>
>> Has a pretty predictable behavior across platforms.  This would save it from
>> being overwritten every time you updated the package.
>>
>> Hope this helps!
>>
>> -Charlie
> 
> I also was about to suggest putting the code in a package because then 
> you have system.file available, which I presume works across platforms:
> 
>  > system.file("RunTests.R",package="DistributionUtils")
> [1] "/usr/local/david/R/DistributionUtils"
>  > system.file("unitTests/runTests.R",package="DistributionUtils")
> [1] "/usr/local/david/R/DistributionUtils/unitTests/runTests.R"


Sorry, I copied that incorrectly. Should be:

 > system.file(package="DistributionUtils")
[1] "/usr/local/david/R/DistributionUtils"
 > system.file("unitTests/runTests.R",package="DistributionUtils")
[1] "/usr/local/david/R/DistributionUtils/unitTests/runTests.R"

David Scott
_________________________________________________________________
David Scott	Department of Statistics
		The University of Auckland, PB 92019
		Auckland 1142,    NEW ZEALAND
Phone: +64 9 923 5055, or +64 9 373 7599 ext 85055
Email:	d.scott at auckland.ac.nz,  Fax: +64 9 373 7018

Director of Consulting, Department of Statistics