[R] Installing package for i386

Sarah Hawley sarah at canaryfoundation.org
Tue Aug 4 01:46:50 CEST 2009

Hi Mark, 
Thanks for the suggestion and for setting me straight about my operating
system. I will contact my collaborator.

On 8/3/09 1:59 PM, "Marc Schwartz" <marc_schwartz at me.com> wrote:

> On Aug 3, 2009, at 3:01 PM, Sarah Hawley wrote:
>> I am trying to install a package from a collaborator that was given
>> to me as
>> a zip file.
>> I'm running R 2.9.0 using Mac OSX 10.5.7
>> [R.app GUI 1.28 (5395) i386-apple-darwin8.11.1]
>> foo.zip contains all the usual package directories (man, R, R-ex, etc)
>> After unzipping foo, I used the following commands.
>> $ R CMD INSTALL foo
>> * Installing to library Œ/Library/Frameworks/R.framework/Resources/
>> library¹
>> * Installing *binary* package Œfoo¹ ...
>> * DONE (foo)
>> I open the R GUI and type
>>> library(foo)
>> Error: package 'flsa' is not installed for 'arch=i386'
>> I assume this error has something to do with my Linux architecture
>> but I
>> have no idea what is really going on.
>> Any suggestions would be appreciated.
>> Thank you.
>> Sarah
> Sarah,
> It looks like your colleague gave you a Windows binary version of the
> package. You cannot just unzip it and install it as such. The ZIP file
> would be the result of running 'R CMD build foo' on the package source
> tree on Windows, which does more than just archive the files in the
> package tree.
> This has nothing to do with Linux BTW, as Mac OSX is a BSD Unix
> derivative, not Linux, albeit they are cousins of a sort...  :-)
> Can your colleague provide you with just the actual source files in a
> tar or zip file without running R CMD build foo? If so, you can at
> least transfer the source files to your Mac and possibly build the
> package there for installation. This process will also be less
> complicated if the package only contains R code and no C or FORTRAN
> code.
> HTH,
> Marc Schwartz

Sarah Hawley
Data Analyst
Canary Foundation
sarah at canaryfoundation.org

Stopping cancer early...
the best possible investment!

More information about the R-help mailing list