Skip to content

R.home() of R 2.8.1 Mac OS X is wrong (PR#13494)

4 messages · andreas.fischlin at env.ethz.ch, Simon Urbanek, Brian Ripley +1 more

Original
andreas.fischlin at env.ethz.ch
# 
Full_Name: Andreas Fischlin
Version: 2.8.1
OS: Mac OS X 10.5.6
Submission from: (NULL) (84.75.178.229)


R.home() returns "/Library/Frameworks/R.framework/Resources" although there is
no such directory present. Inspecting the actual directory structure of the R
installation indicates that the correct value R.home() should contain is:
'/Library/Frameworks/R.framework/Versions/2.8/Resources'

Detected by af, ETH Zurich / 1.Feb.2009 using R 2.8.1 under Mac OS X 10.5.6
Simon Urbanek
#
On Feb 1, 2009, at 1:50 PM, andreas.fischlin at env.ethz.ch wrote:

            

      
      
      
    

        
Then your installation is broken. You should re-install R to fix your  
installation.

In R 2.8.1:
fino:~$ file /Library/Frameworks/R.framework/Resources
/Library/Frameworks/R.framework/Resources: directory

This is true for all R installations from CRAN as well as all custom  
installations with framework enabled (default on OS X).

Cheers,
S

            

      
      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Brian Ripley
    

    

      
      #
    

  


  

      
It is present on my system (R reinstalled yesterday), so it looks like 
you broke yours.

More precisely, there is a link at 
/Library/Frameworks/R.framework/Resources to the correct place.

How did you think R could possibly work with this broken?  R.home() is 
used all over the place in R's own scripts.

        
On Sun, 1 Feb 2009, andreas.fischlin at env.ethz.ch wrote:

        

            

      
      
      
    

        

      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Andreas Fischlin
    

    

      
      #
    

  


  

      
Thank you all for the replies. However, I figured it now out, there is 
nothing broken. The reason I got confused is only because there are 
symbolic links used extensively and some tests failed in my script as a 
consequence (see below).

My installation is fine and looks in fact like this:

afischli$ cd /Library/Frameworks/R.framework
afischli$ ls -l
total 40
lrwxr-xr-x  1 root  admin   24 Jan 13 15:24 Headers -> 
Versions/Current/Headers
lrwxr-xr-x  1 root  admin   30 Jan 13 15:24 Libraries -> 
Versions/Current/Resources/lib
lrwxr-xr-x  1 root  admin   31 Jan 13 15:24 PrivateHeaders -> 
Versions/Current/PrivateHeaders
lrwxr-xr-x  1 root  admin   18 Jan 13 15:24 R -> Versions/Current/R
lrwxr-xr-x  1 root  admin   26 Jan 13 15:24 Resources -> 
Versions/Current/Resources
drwxrwxr-x  7 root  admin  238 Feb  1 20:17 Versions
afischli$ cd Versions
afischli$ ls -l
total 8
drwxrwxr-x  6 root  admin  204 May 24  2007 2.5
drwxrwxr-x  6 root  admin  204 Nov 30  2007 2.6
drwxrwxr-x  6 root  admin  204 Jan 13 15:24 2.8
lrwxr-xr-x  1 root  admin    3 Jan 13 15:24 Current -> 2.8

This means there is nothing broken. The consequence is that R commands 
such as:

 > R.home()
[1] "/Library/Frameworks/R.framework/Resources"
 > setwd(R.home())
 > getwd()
[1] "/Library/Frameworks/R.framework/Versions/2.8/Resources"

give differing answers, which means that a test such as

setwd(R.home())
if (getwd() == R.home()) {

always fails, although the setwd command was actually successful. Guess 
this all explains the problem and sorry for the bandwidth waste ;-)

Regards,
Andreas Fischlin

________________________________________________________________________
ETH Zurich
Prof. Dr. Andreas Fischlin
Systems Ecology - Institute of Integrative Biology
CHN E 21.1
Universitaetstrasse 16
8092 Zurich
SWITZERLAND

andreas.fischlin at env.ethz.ch
www.sysecol.ethz.ch

+41 44 633-6090 phone
+41 44 633-1136 fax

             Make it as simple as possible, but distrust it!
________________________________________________________________________

        
Prof Brian Ripley wrote: