Hi Monnie -- I'm hoping that someone with Mac experience will provide
specific guidance. The error message says that the file libgvc.3.dylib
is not found. This is likely because it is installed somewhere that
Rgraphviz does not know about. I don't know if this is an issue with
the way graphviz was installed on your Mac (where is libgvc.3.dylib?)
or how Rgraphviz looks for this package (what does
Sys.getenv("LD_LIBRARY_PATH") say from within R?).
The 'README' file I was referring to was the README file in the
Rgraphviz package; visit
http://bioconductor.org/packages/2.3/bioc/html/Rgraphviz.html
download and unpack the 'Package source' link to view.
Martin
"McGee, Monnie" <mmcgee at="" mail.smu.edu=""> writes:
> Dear Martin,
>
> Thank you. What are the steps for Macintosh? I've copied in the
error and my sessionInfo below.
>
> As far as I can tell, the version of Graphviz that I downloaded for
the Mac (Version 2.20 (2.20.3)) does not have a README file. Do I
need to download that from the Graphviz website, also?
>
> Thanks,
> Monnie
>
> ##### Error I received ######
>> library(Rgraphviz)
> Error in dyn.load(file, DLLpath = DLLpath, ...) :
> unable to load shared library '/Library/Frameworks/R.framework/Res
ources/library/Rgraphviz/libs/i386/Rgraphviz.so':
> dlopen(/Library/Frameworks/R.framework/Resources/library/Rgraphviz
/libs/i386/Rgraphviz.so, 6): Library not loaded:
/usr/local/lib/libgvc.3.dylib
> Referenced from: /Library/Frameworks/R.framework/Resources/library
/Rgraphviz/libs/i386/Rgraphviz.so
> Reason: image not found
> Error : .onLoad failed in 'loadNamespace' for 'Rgraphviz'
> Error: package/namespace load failed for 'Rgraphviz'
>
> ### Session Info #####
>> sessionInfo()
> R version 2.7.2 (2008-08-25)
> i386-apple-darwin8.11.1
>
> locale:
> en_US.UTF-8/en_US.UTF-8/C/C/en_US.UTF-8/en_US.UTF-8
>
> attached base packages:
> [1] tools grid stats graphics grDevices utils
datasets methods
> [9] base
>
> other attached packages:
> [1] Rintact_1.2.0 hypergraph_1.12.0 XML_1.96-0
yeastExpData_0.9.11
> [5] RBGL_1.16.0 RColorBrewer_1.0-2 Biobase_2.0.1
graph_1.18.1
>
> loaded via a namespace (and not attached):
> [1] cluster_1.11.11
>>
> ________________________________________
> From: Martin Morgan [mtmorgan at fhcrc.org]
> Sent: Monday, November 17, 2008 5:10 PM
> To: Raffaele calogero
> Cc: McGee, Monnie; bioconductor at stat.math.ethz.ch
> Subject: Re: [BioC] (no subject)
>
> Raffaele calogero wrote:
>> On Mon, Nov 17, 2008 at 4:59 PM, McGee, Monnie <mmcgee at="" mail.smu.edu=""> wrote:
>> Dear Monnie,
>> thank you for your answer.
>> I Rgraphviz is always somehow problematic on windows computers.
>> I do not think I have to download Graphviz since it was never
>> necessary before on packages compiled for windows. The strange
point
>> is that I can load and use without problems the Bioconductor 2.1
>> Rgraphviz version.
>
> Windows versions of Rgraphviz working with R 2.7.* or 2.8.* require
that
> you have graphviz installed. It must be a specific version of
Graphviz.
> The README file for the CURRENT version (for use with R-2.8.*) of
> Rgraphviz says
>
>> Windows
>> =======
>>
>> Nov 4, 2008
>>
>> Here are the steps to build Rgraphviz on Windows:
>>
>> o. install graphviz ***2.20.3*** Windows ***binary*** distribution
from
>>
http://www.graphviz.org/pub/graphviz/stable/windows/graphviz-
2.20.3.msi
>> , say, in dir $DIR. We use the following as an example:
>> C:\Graphviz-2.21
>> . If you install it in other dir, just use your dir
correspondingly.
>>
>> o. set the following Windows environment variables accordingly
>> (control panel -> systems -> Advanced -> Environment Variables
):
>> (a) create new user variables:
>> GRAPHVIZ_INSTALL_DIR (e.g., C:\Graphviz-2.21)
>> GRAPHVIZ_INSTALL_MAJOR (e.g., 2)
>> GRAPHVIZ_INSTALL_MINOR (e.g., 20)
>>
>> Notice the way "\/" are used for the paths above.
>>
>> (b) add to user variable 'path': $GRAPHVIZ_INSTALL_DIR/bin
>> e.g., C:\Graphviz-2.21\bin
>>
>> o. run "R CMD INSTALL ...Rgraphviz" as usual
>
> The specific version of graphviz is different for R-2.7.*.
>
> Martin
>
>> Cheers
>> Raffaele
>>
>>> Dear Raffaele,
>>>
>>> I have had the same problem on my Macintosh computer. I sent an
e-mail to the list last Thursday, but I have not yet received a reply.
>>>
>>> One question for you - have you downloaded Graphviz? Rgraphviz
will not work without Graphviz. The website is www.graphiviz.org.
>>>
>>> I downloaded Graphiviz, and it didn't solve my problem, but it may
solve yours.
>>>
>>> Good Luck!
>>> Monnie
>>>
>>>
>>> Message: 1
>>> Date: Sun, 16 Nov 2008 16:39:14 +0100
>>> From: rcaloger <raffaele.calogero at="" gmail.com="">
>>> Subject: [BioC] Rgraphviz error under Windows
>>> To: bioconductor at stat.math.ethz.ch
>>> Message-ID: <49203EA2.7030403 at unito.it>
>>> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>>>
>>> HI,
>>> I installed Rgraphviz_1.20.2.zip without error on Windows Vista 32
bits
>>> Centrino duo with 2Gb RAM.
>>> But when I tryed to load the library:
>>> library(Rgraphviz)
>>> I got the following error:
>>> Error in get(hookname, envir = env, inherits = FALSE) :
>>> cannot allocate memory block of size 2.2 Gb
>>> Error: package/namespace load failed for 'Rgraphviz'
>>>
>>> This happens on an clean R session without anything loaded.
>>> When I tryed to install the Rgraphviz from Bioconductor 2.2
>>> Rgraphviz_1.18.1.zip, instllation succeded but the loading of the
>>> library produced an other error:
>>> > library(Rgraphviz)
>>> Loading required package: graph
>>> Loading required package: grid
>>> Error in inDL(x, as.logical(local), as.logical(now), ...) :
>>> unable to load shared library
>>> 'C:/R/R-2.8.0/library/Rgraphviz/libs/Rgraphviz.dll':
>>> LoadLibrary failure: Unable to find the specified module.
>>>
>>> Error : .onLoad failed in 'loadNamespace' for 'Rgraphviz'
>>> Error: package/namespace load failed for 'Rgraphviz'
>>>
>>> Loading of the Rgraphviz lib was instead successful using
>>> Rgraphviz_1.16.0.zip from Bioconductor 2.1
>>>
>>> Any suggestion how to solve Rgraphviz loading problem?
>>> Raffaele
>>
>>
>>
>
>
> --
> Martin Morgan
> Computational Biology / Fred Hutchinson Cancer Research Center
> 1100 Fairview Ave. N.
> PO Box 19024 Seattle, WA 98109
>
> Location: Arnold Building M2 B169
> Phone: (206) 667-2793
--
Martin Morgan
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109
Location: Arnold Building M2 B169
Phone: (206) 667-2793