[Phylobase-devl] Urgent! problems with installing
Marguerite Butler
mbutler at hawaii.edu
Thu Jul 17 10:29:08 CEST 2008
Hi Everyone,
Thanks for all your replies earlier. I just wanted to report that the
problem with read.nexus was a bug in ape that Emmanuel fixed in the
latest release. It was a bug that only occurred in a single-tree file
with no translation block.
I figured out my problems with phylobase on R 2.6.2. I had previously
upgraded to 2.7 but then downgraded again. However, the 2.7 directory
was still present on my harddrive, it's just that the current version
was pointing to the 2.6 folder (I guess R keeps track of the current
version by using an alias called current pointing to the appropriate
folder). I guess the package installer just looks at the highest
version number you have and not what is pointed to by "current". Once
I moved the 2.7 folder to another location, all was fine with
installing phylobase from source on 2.6.2.
Marguerite
On Jul 2, 2008, at 6:09 AM, Marguerite Butler wrote:
> Hi Francois, Brian, and Ben,
>
> Thank you for looking into this. But to the ape developers: has the
> code for read.nexus changed recently? Or is this a problem with the
> R version?
>
> I installed ape from binaries each time, which is ape 2.1-3. I got
> the bug with it installed in R 2.7.1 but not R 2.6.2.
>
> I downloaded ape 2.1-3 and ape 2.2 sources, and did a diff on the
> read.nexus.R files, and got no differences. I'm worried about what
> else will come up if I use R 2.7 for the comparative methods course.
>
> Marguerite
>
>
> On Jul 2, 2008, at 5:48 AM, François Michonneau wrote:
>
>> Hi Marguerite
>>
>> I think I found the bug. I looked at that very quickly so I might
>> be wrong.
>>
>> It comes from the function read.nexus. Towards the end there is:
>>
>> if (Ntree == 1) {
>> trees <- trees[[1]]
>> #trees$tip.label <- TRANS[, 2]
>> }
>> else {
>> if (!is.null(tree.names))
>> names(trees) <- tree.names
>> if (translation)
>> attr(trees, "TipLabel") <- TRANS[, 2]
>> class(trees) <- "multiPhylo"
>> }
>> if (length(grep("[\\/]", file)) == 1)
>> file <- paste(getwd(), file, sep = "/")
>> attr(trees, "origin") <- file
>> trees
>>
>>
>> However, the variable TRANS is defined only if the word
>> "translate" is present (look at the beginning of read.nexus) in
>> the nexus file. Thus the error message. I haven't spent enough
>> time to try to figure out what this line was doing specifically,
>> but if this line is commented then there is no error message and
>> it still works.
>>
>> Cheers,
>> François
>>
>> Marguerite Butler wrote:
>>> Hi Ben,
>>> Thanks for the quick reply!
>>>>
>>>> No problems here with 2.7.0.
>>>> Can you send a read.nexus example for me to try?
>>>>
>>> Here is the nexus file that read.nexus couldn't read under R
>>> 2.7.1 yesterday (this was a problem both for macs and pc's on
>>> several people's computers -- I had a small group of students try
>>> it). When I downgraded it read in fine.
>>> --------------------------------------------------------------------
>>> ----
>>> It would be great if you could give it a try.
>>>>
>>>>> The binaries were available on the r-forge site for a while,
>>>>> but not anymore.
>>>>
>>>> It looks (?) like they're there -- I even downloaded the file
>>>> and it seems to have the right bits in it, but I can't tell
>>>> without trying to install it ...
>>>>
>>> Oops! Sorry you are right, the binaries are there. I guess I got
>>> myself confused because it is not passing the check. However, it
>>> is not possible to install the binaries unless you have 2.7. The
>>> real problem is that I can't install from source in 2.6.2.
>>>> But my problem is that I can't even install from source. When I
>>>>> did R CMD check on the package, I get one warning:
>>>>> ====
>>>>> * checking if this is a source package ... WARNING
>>>>> Subdirectory 'pkg/src' contains object files.
>>>>> ====
>>>>> I found "phylobase.so" in pkg/src. Is this supposed to be
>>>>> there? It might be the culprit because the error I get when I
>>>>> try to load phylobase is:
>>>>> > require(phylobase)
>>>>> Loading required package: phylobase
>>>>> Error in dyn.load(file, ...) :
>>>>> unable to load shared library '/Library/Frameworks/
>>>>> R.framework/Resources/library/phylobase/libs/i386/
>>>>> phylobase.so': dlopen(/Library/Frameworks/R.framework/
>>>>> Resources/library/phylobase/libs/i386/phylobase.so, 6): Library
>>>>> not loaded: /Library/Frameworks/R.framework/Versions/2.7/
>>>>> Resources/lib/libR.dylib
>>>>> Referenced from: /Library/Frameworks/R.framework/Resources/
>>>>> library/phylobase/libs/i386/phylobase.so Reason: image not found
>>>>> Which doesn't make a whole lot of sense since I'm installing
>>>>> from source on R 2.6.2.
>>>>
>>>> It seems it's still looking for a 2.7 library?
>>> Yes, that is what it looks like to me. Why would it be doing
>>> that? I tried trashing the phylobase directory in the location on
>>> my computer where R is installed, which is here: /Library/
>>> Frameworks/R.framework/Resources/library/phylobase
>>> before re-installing, in case there was something in there from a
>>> previous installation with 2.7 still lurking around, but it
>>> didn't seem to help. Is there anything else I should do for a
>>> "clean" install of phylobase?
>>>>
>>>>> So I went back and removed the "phylobase.so" from the pkg/src
>>>>> directory (the version I got from the svn repository --
>>>>> revision 192), and tried rebuilding. It passed the check this
>>>>> time, so I installed. But when I try loading, I still get the
>>>>> same error above from my require(phylobase) command??
>>>>> Has anyone else had problems?
>>>>> I don't think it's entirely a 2.6.2 issue because I don't think
>>>>> it's working for the 2.7.X versions either, otherwise wouldn't
>>>>> there be binaries available?
>>>>> I am preparing for the NESCENT course, so this is pretty
>>>>> urgent. Any help would be greatly appreciated.
>>>>> Marguerite
>>>>
>>>> I will do anything I can to help but am a little restricted
>>>> since it's working fine for me and I don't have a Mac ...
>>>>
>>> If it turns out that we can only use ape with 2.6.2, and
>>> phylobase with 2.7, is there a way to switch between R versions
>>> on the same computer (of course in different R sessions)? I know
>>> that the old versions are still there when you install a newer
>>> version. Has any of the other mac users had problems? Thanks,
>>> Marguerite
>>> ____________________________________________
>>> Marguerite A. Butler
>>> Department of Zoology
>>> University of Hawaii
>>> 2538 McCarthy Mall, Edmondson 259
>>> Honolulu, HI 96822
>>> Phone: 808-956-4713
>>> FAX: 808-956-9812
>>> Dept: 808-956-8617
>>> http://www2.hawaii.edu/~mbutler
>>> http://www.hawaii.edu/zoology/
>>> --------------------------------------------------------------------
>>> ----
>>> _______________________________________________
>>> Phylobase-devl mailing list
>>> Phylobase-devl at lists.r-forge.r-project.org
>>> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/
>>> phylobase-devl
>
> ____________________________________________
> Marguerite A. Butler
> Department of Zoology
> University of Hawaii
> 2538 McCarthy Mall, Edmondson 259
> Honolulu, HI 96822
>
> Phone: 808-956-4713
> FAX: 808-956-9812
> Dept: 808-956-8617
> http://www2.hawaii.edu/~mbutler
> http://www.hawaii.edu/zoology/
>
> _______________________________________________
> Phylobase-devl mailing list
> Phylobase-devl at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/
> phylobase-devl
____________________________________________
Marguerite A. Butler
Department of Zoology
University of Hawaii
2538 McCarthy Mall, Edmondson 259
Honolulu, HI 96822
Phone: 808-956-4713
FAX: 808-956-9812
Dept: 808-956-8617
http://www2.hawaii.edu/~mbutler
http://www.hawaii.edu/zoology/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.r-forge.r-project.org/pipermail/phylobase-devl/attachments/20080716/b15df62a/attachment-0001.htm
More information about the Phylobase-devl
mailing list