[R] Failure loading library into second R 2.3.1 session on Windows XP

[Uwe Ligges]

Talbot Katz wrote:
> Hi Uwe.
> 
> Thank you so much for responding!  I guess I wasn't entirely clear about 
> the problem.  If I make the mistake of trying to install a package from 
> CRAN in a second session after I've already installed it in a previous 
> session, it won't install in the second session, and even if I close the 
> second session and open a subsequent newer session, it won't install in 
> that one either.  At least, I can't figure out how to do it, because 
> it's no longer in the "Load packages..." menu, and if I load it from 
> CRAN, I get that funny error message:
>     Warning: cannot remove prior installation of package 'corpcor'


> Now, after having gone through this, I know enough not to reload a 
> package from CRAN.  But it appears that the only ways to solve the 
> problem, if it occurs, are pretty drastic, either reboot the machine 
> (which is what I did) or reinstall R (which seems to be what you're 
> suggesting?).  I was hoping that there might be a better alternative, or 
> at least that the development team might look into this issue for future 
> releases.  This doesn't affect every package, but I've seen it in the 
> first two packages I tried it with.

No, you just have to close all running R processes, restart R and 
reinstall the package (without having it loaded by any R process). There 
is no reason to reboot the machine.

Please distinguish between installing and loading a package.

Uwe Ligges



> --  TMK  --
> 212-460-5430    home
> 917-656-5351    cell
> 
> 
> 
>> From: Uwe Ligges <ligges at statistik.uni-dortmund.de>
>> To: Talbot Katz <topkatz at msn.com>
>> CC: r-help at stat.math.ethz.ch
>> Subject: Re: [R] Failure loading library into second R 2.3.1 session 
>> on Windows XP
>> Date: Fri, 29 Dec 2006 23:05:15 +0100
>>
>> You can only expect that update / reinstall a ***package*** works if 
>> you have not yet loaded it into your R session.
>> Hence close R, start it without loading the relevant package and then 
>> update/reinstall.
>>
>> Best,
>> Uwe Ligges
>>
>>
>> Talbot Katz wrote:
>>> Hi.
>>>
>>> I am using R 2.3.1 on Windows XP.  I had installed a library package 
>>> into my first session and wanted the same package in my second 
>>> session, so I went out to the CRAN mirror and tried to install the 
>>> package, and got the following message:
>>>
>>> *********************************************************************
>>>
>>>> utils:::menuInstallPkgs()
>>> trying URL 
>>> 'http://cran.ssds.ucdavis.edu/bin/windows/contrib/2.3/corpcor_1.4.4.zip'
>>> Content type 'application/zip' length 133068 bytes
>>> opened URL
>>> downloaded 129Kb
>>>
>>> package 'corpcor' successfully unpacked and MD5 sums checked
>>> Warning: cannot remove prior installation of package 'corpcor'
>>>
>>> The downloaded packages are in
>>>         C:\Documents and Settings\Talbot\Local 
>>> Settings\Temp\RtmplCxarb\downloaded_packages
>>> updating HTML package descriptions
>>>> library(corpcor)
>>> Error in library(corpcor) : there is no package called 'corpcor'
>>>
>>> *********************************************************************
>>>
>>>
>>> After rebooting my machine, I dug into this a little further.  Upon 
>>> installing a package from a CRAN mirror, it seems to stay on my hard 
>>> drive, and I can load it in subsequent sessions from the "Load 
>>> package..." menu without going back to get it from a CRAN mirror.  
>>> However, if I do happen to retrieve it again from a CRAN mirror, it 
>>> appears that may corrupt the version that was saved, and it no longer 
>>> will be available from the "Load package..." menu.  A reboot and 
>>> re-retrieval of the package makes it available again; I don't know 
>>> whether there's any less drastic solution.
>>>
>>> This behavior doesn't occur with every package, but I have 
>>> experienced it with two different packages (corpcor and copula), so 
>>> there seems to be something going on.  I didn't see anything in the 
>>> FAQ page about this, I wonder if anyone can tell me more about this 
>>> issue.
>>>
>>> Thanks!
>>>
>>>
>>> --  TMK  --
>>> 212-460-5430    home
>>> 917-656-5351    cell
>>>
>>> ______________________________________________
>>> R-help at stat.math.ethz.ch 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.
> 
>