[R-SIG-Mac] arules installation

Thu Dec 8 14:57:40 CET 2011

On Dec 8, 2011, at 5:16 AM, Thomas Chesney wrote:

> Hey it worked! Thank you *very* much. I installed XCode from my Mac  
> DVD, then downloaded the arules source file and ran R CMD INSTALL / 
> Users/thomas/Documents/arules_1.0-7.tar.gz in the terminal.
>
> So should I always use the source file of the pakages in the future?

I wouldn't. I used to do so but the system is very stable in my  
experience over the last several years. Unless you enjoy the thrill of  
seeing pages of compiler messages scrolling across you screen,  (and I  
will admit to sometimes being strangely thrilled by that experience at  
times) there isn't much point.

> How come this worked but downloading the binary didn't? The binary  
> file is somehow tied to a version of R whereas the source file isn't?

To some extent that may be true. Major version upgrades to R sometimes  
add requirements on packages or provide new functions. Source code  
from former years may also fail to compile with newer versions of R.

> Has the binary file been complied already whereas my CMD INSTALL  
> command compiled C code on my machine?

Yes. Mac binaries are *.tgz files whereas *.tar.gz files are source.  
Source files may or may not require compilation of C/Fortran code. Not  
all packages have such. Some packages are pure R and installation  
would succeed even without the Xcode resources.

> But then why would my compiled C code be different from the compiled  
> C code in the binary file?

Not sure what you mean here. Are you asking about some specific  
instance?

-- 
David.

>
> Anyway, important thing is it seems to have worked, but if someone  
> could advise on whether I should use source files in the future that  
> would be really helpful.
>
> Thomas
> ________________________________________
> From: David Winsemius [dwinsemius at comcast.net]
> Sent: Thursday, December 08, 2011 12:58 AM
> To: R.Michael Weylandt
> Cc: Thomas Chesney; r-sig-mac at r-project.org
> Subject: Re: [R-SIG-Mac] arules installation
>
> On Dec 7, 2011, at 4:35 PM, R. Michael Weylandt wrote:
>
>> Theoretically, but I've always built packages from Terminal so I  
>> can't
>> promise it. Also, there's an outside chance you'll need a Fortran
>> compiler too, but I haven't looked at the source files to confirm.
>>
>> I'd also suggest you try your CRAN mirror one more time and make sure
>> you are getting the 2.13 version: for instance, I used this link:
>>
>> http://lib.stat.cmu.edu/R/CRAN/bin/macosx/leopard/contrib/2.13/arules_1.0-7.tgz'
>>
>> I believe that the link shown on the CRAN page will automatically  
>> give
>> you the current release (2.14) build.
>>
>> Michael
>>
>> 2011/12/7 Thomas Chesney <Thomas.Chesney at nottingham.ac.uk>:
>>> Michael - thanks for the reply. So if I install my XCode C complier,
>
> You would want all of the Xcode to be installed including the Fortran
> complier and other support.
>
>>> then download the latest arules source file which is:
>>> arules_1.0-7.tar.gz and then run: install.packages("/Users/thomas/
>>> Documents/arules_1.0-7.tar.gz", repos=NULL, type="source") it will
>>> just work? (Why will that work with the latest version but the
>>> latest .tgz won't?)
>
> It should succeed that way if Xocde is properly installed. Why didn't
> you try that?
>
> AND: Why are you staying with 2.13? Why not update your R version?
>
>>>
>>> What I usually do to install a package is download the .tgz file,
>>> double click to unzip it, then copy the resulting folder into
>>> Library/Frameworks/R.Framework/Resources/Library.
>
> That is only going to work for  packages that are all-R and no C/ 
> Fortran
>
>>> I could install the latest R and then do that for arules, but then
>>> I'd have to do the same for dozens of other packages (or could I
>>> just copy them to another folder, update R, and then copy them to
>>> R.Framework/Resources/Library saying no to any replace file popups?)
>
> The usual recommendation is after updating R , which creates a new
> Resources folder with ./library/, to copy/paste the contents of the
> 2.13 ../library/ into the new one and then at the R console to  
> execute:
>
> update.packages(checkBuilt=TRUE) # and probably also include the
> ask=FALSE option
>
> I usually forget that ask option and then have to keep typing Y<cr>.
>
> Best;
> David.
>
>
>>>
>>> Thomas
>>>
>>> ________________________________________
>>> From: R. Michael Weylandt <michael.weylandt at gmail.com> [michael.weylandt at gmail.com
>>> ]
>>> Sent: Wednesday, December 07, 2011 1:36 PM
>>> To: Thomas Chesney
>>> Cc: r-sig-mac at r-project.org
>>> Subject: Re: [R-SIG-Mac] arules installation
>>>
>>> You need a C compiler: the usual recommendation is XCode which you
>>> can get off your OS X install DVD or download from Apple Developer
>>> pages.
>>>
>>> Out of curiosity, how did you download the first time? I can still
>>> get 2.13.x packages from CRAN though I haven't tried with arules
>>> specifically.
>>>
>>> Michael
>>>
>>> On Dec 7, 2011, at 6:05 AM, Thomas Chesney <Thomas.Chesney at nottingham.ac.uk
>>>> wrote:
>>>
>>>> I sent the following and got the following reply to the R-Help,
>>>> and then was told I should have sent it here.
>>>>
>>>> On 07.12.2011 10:54, Thomas Chesney wrote:
>>>>> I'm using R version 2.13.0 (2011-04-13) on Mac OS X and I get the
>>>>> following error message with library(arules):
>>>>>
>>>>> Loading required package: Matrix
>>>>> Loading required package: lattice
>>>>>
>>>>> Attaching package: 'Matrix'
>>>>>
>>>>> The following object(s) are masked from 'package:base':
>>>>>
>>>>>   det
>>>>>
>>>>> Error in dyn.load(file, DLLpath = DLLpath, ...) :
>>>>> unable to load shared object '/Library/Frameworks/R.framework/
>>>>> Versions/2.13/Resources/library/arules/libs/i386/arules.so':
>>>>> dlopen(/Library/Frameworks/R.framework/Versions/2.13/Resources/
>>>>> library/arules/libs/i386/arules.so, 6): Library not loaded: /
>>>>> Library/Frameworks/R.framework/Versions/2.14/Resources/lib/
>>>>> libR.dylib
>>>>> Referenced from: /Library/Frameworks/R.framework/Versions/2.13/
>>>>> Resources/library/arules/libs/i386/arules.so
>>>>> Reason: image not found
>>>>> In addition: Warning message:
>>>>> package 'arules' was built under R version 2.14.0
>>>>> Error: package/namespace load failed for 'arules'
>>>>
>>>> Install an arules binary that was built for your version of R
>>>> (reather
>>>> than a later one) or install from sources.
>>>>
>>>> Uwe Ligges
>>>>
>>>>>
>>>>> The arules CRAN page says: Depends:   R (≥ 2.11.0), stats,
>>>>> methods, Matrix (≥ 0.999375-38), all of which I have - anyone
>>>>> know what the problem is? I'm really relunctant to update R as
>>>>> its Package Installer is blocked somehow by our Firewall and
>>>>> updating all the packages I use is a major hassle.
>>>>>
>>>>> Thank you (as always!)
>>>>>
>>>>> Thomas
>>>>>
>>>>> PS on a related note, are the package installer port settings
>>>>> available somewhere and I'll send them to IT, ask them to unblock
>>>>> it?
>>>>
>>>>
>>>> -------------------------------------
>>>>
>>>> I tried installing from source as Prof Ligges suggested (as
>>>> explained here: http://stackoverflow.com/questions/1474081/how-do-i-install-an-r-package-from-source)
>>>> but got the following error:
>>>>
>>>> install.packages("/Users/thomas/Documents/arules_1.0-7.tar.gz",
>>>> repos=NULL, type="source")
>>>> * installing *source* package 'arules' ...
>>>> ** libs
>>>> *** arch - i386
>>>> sh: make: command not found
>>>> ERROR: compilation failed for package 'arules'
>>>> * removing '/Library/Frameworks/R.framework/Versions/2.13/
>>>> Resources/library/arules'
>>>> Warning message:
>>>> In install.packages("/Users/thomas/Documents/
>>>> arules_1.0-7.tar.gz",  :
>>>> installation of package '/Users/thomas/Documents/
>>>> arules_1.0-7.tar.gz' had non-zero exit status
>>>>
>>>> Why would installing from source make any difference? Prof Ripley
>>>> suggested I use arules for R.13.0, which I would if I could find  
>>>> it.
>>>>
>>>> Has anyone got any advice?
>>>>
>>>> Thank you,
>>>>
>>>> Thomas
>>>> This message and any attachment are intended solely for the
>>>> addressee and may contain confidential information. If you have
>>>> received this message in error, please send it back to me, and
>>>> immediately delete it.   Please do not use, copy or disclose the
>>>> information contained in this message or in any attachment.  Any
>>>> views or opinions expressed by the author of this email do not
>>>> necessarily reflect the views of the University of Nottingham.
>>>>
>>>> This message has been checked for viruses but the contents of an
>>>> attachment
>>>> may still contain software viruses which could damage your
>>>> computer system:
>>>> you are advised to perform your own checks. Email communications
>>>> with the
>>>> University of Nottingham may be monitored as permitted by UK
>>>> legislation.
>>>> _______________________________________________
>>>> R-SIG-Mac mailing list
>>>> R-SIG-Mac at r-project.org
>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>
>> _______________________________________________
>> R-SIG-Mac mailing list
>> R-SIG-Mac at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>
> David Winsemius, MD
> West Hartford, CT
>

David Winsemius, MD
West Hartford, CT