[R] Upgrading R means I lose my packages

Charilaos Skiadas cskiadas at gmail.com
Thu Aug 28 05:00:35 CEST 2008


On Aug 27, 2008, at 10:40 PM, Rolf Turner wrote:

>
> On 28/08/2008, at 2:02 PM, James Milks wrote:
>
>> The title says it all.  Does anyone know of a way to save your  
>> packages when you upgrade to a new version of R?  This may seem  
>> petty, but I'm accumulating enough packages that having to  
>> download and install each of them anew every time I install a new  
>> version of R is rather of a pain.  Ideally, I would like the new  
>> version of R to recognize the packages I've installed on the  
>> previous version without needing to reinstall the packages.  Is  
>> that possible?
>>
>> My system: Mac OS 10.5.4.
>> Current R version: 2.7.1
>
> Mac OS moves in mysterious ways, but apparently your installation  
> moves in more mysterious
> ways than most.
>
> I also (by necessity, not by choice) run Mac OS.  But I certainly  
> don't lose my
> packages when I update R.  The new version of R certainly  
> ``recognizes'' the packages
> that I have installed.  No action required.
>
> There may be something funny about *where* you have your packages  
> installed, and
> what environment variables you have set.
>
> To answer your question ``Is that possible?'' --- Yes.  Not just  
> possible,
> but universal.  Except, it would seem, in your case.  What have you  
> done
> to offend the gods? :-)
>

Actually have had the same problem as James. By default, unless I'm  
mistaken, R will save installed packages within the "R.framework"  
framework (system-wide installation). This framework gets completely  
replaced when a new version is installed. In my system, the location  
of these packages is:
/Library/Frameworks/R.framework/Resources/library

So unless I am mistaken you have to take some action to prevent  
packages from being installed there. I do hope I am wrong.

> 	cheers,
>
> 		Rolf Turner
>


Haris Skiadas
Department of Mathematics and Computer Science
Hanover College



More information about the R-help mailing list