IF you are using a CRAN binary version of R, it is strange how fink is
involved at all. You must have a messed up system
My best guess is that DYLD_LIBRARY_PATH or LD_LIBRARY_PATH has been
set (you can do
echo $LD_LIBRARY_PATH
in the terminal). Unset those variables, and proceed.
Kasper
On
CURL_CONFIG=/sw/bin/curl-config
When I was trying to get RCurl working this was suggested as the way to do
it.
So I assume i should not do:
export DYLD_LIBRARY_PATH=/sw/lib:/usr/local/lib:
On Mon, Nov 15, 2010 at 3:38 PM, Kasper Daniel Hansen
kasperdanielhan...@gmail.com wrote:
On Mon, Nov
Ricardo
The problem is that the place where R puts the Sweave file is not part
of the search path for your tex installation. You can fix this by
adding the path as you would add any other path, or you can use the
following in your .Rprofile which will put an absolute path inside
your generated
When you install from the GUi you can either do a system wide install
(in /Library/...) or you can do a user-level install (in
~/Library/...).
You can see where R searches for packages by doing
.libPaths()
I expect that the first element will be somewhere in your home dir.
I don't use the GUI
You are right. Essentially you can change compilers and compiler
flags using a set of environment variables. These variables are
common across projects where the build process is managed by automake.
This is especially important if you want to switch between 32 and 64
bit. My build R script
You need to install Xcode.
Kasper
On Mon, Mar 29, 2010 at 4:45 PM, Devin Johnson devin.john...@noaa.gov wrote:
Hi,
I have a problem compiling/installing a package that has fortran source code
in it. I've compiled earlier versions of this package many times in the past
as I've worked on it,
There might be a better way, but on the mac open opens a file as if
you doubleclick on it in Finder, and the system command in R
executes a shell command, so you should be able to do
R system(paste(open, filename))
Kasper
On Wed, Mar 24, 2010 at 7:37 PM, Gi-Mick Wu mick...@mail.mcgill.ca wrote:
not understand this issue.
Best regards
Christian
Kasper Daniel Hansen wrote:
On Jan 23, 2010, at 8:37 AM, cstrato wrote:
Dear Kasper,
Thank you for your suggestion, I did indeed use gcc 4.0.
Since gcc_select does no longer exist on Leopard I had to do:
cd
Hmm, I forgot about --arch. As Simon said, you should be able to do it by
R --arch=x86_64 --no-multiarch CMD INSTALL
My guess is that you need --no-multiarch, but I cannot experiment with this
where I sit.
Kasper
On Jan 26, 2010, at 15:57 PM, Kasper Daniel Hansen wrote:
On Jan 26, 2010
, but in the Tiger section, so I did miss it.
Nevertheless it would still be helpful for Mac users to know how to uninstall
it:
$ tar -tf gfortran.tar | sort -r | (cd /; xargs -p -n 1 rm -d)
Best regards
Christian
Kasper Daniel Hansen wrote:
Remember that Apple has two version of GCC
Remember that Apple has two version of GCC on Leopard: 4.0 and 4.2. You are
using 4.0, you might want to switch to 4.2. In Tiger there used to be
gcc_select (or select_gcc) that let you choose between 3.x and 4.0, I don't
remember if that is still around on Leopard.
The error seems to
Leopard, latest version, using Apples gcc 4.2. SVN version r50716
I get a crash (memory not mapped). The traceback shows grid.Call(L_cap),
followed by .Call(fname, , package = grid)
Kasper
On Dec 13, 2009, at 21:47 PM, Paul Murrell wrote:
Hi
In the *development* version of R, there
Fresh R-devel (as well as R-2.10) on a fresh install of Snow Leopard (not an
upgrade). Using Xcode 3.2.1. Building in a separate directory from the source.
../R-devel-src/configure $OPTIONS
make
sudo make install
The sudo make install fails with
SNIP
mkdir
On Dec 1, 2009, at 11:29 AM, Simon Urbanek wrote:
ls -l /Library/Frameworks/R.framework
ls -l /Library/Frameworks/R.framework/Versions
ls -l /Library/Frameworks/R.framework/Versions/2.11
ls -l /Library/Frameworks/R.framework/Versions/2.11/Resources
This is my current state, after a failed
or something such that it's not a symlink.
If you remove that all should be well. What you have there is definitely not
a clean system ;). I have put a guard about such strange framework setup in
the R-devel, just in case.
Cheers,
Simon
On Dec 1, 2009, at 12:13 , Kasper Daniel Hansen wrote
I think I found the culprit. I installed the devpack before I compiled R, and
looking at the contents of the tarball it does install things in
/Library/Frameworks/R.framework/Resources. That must have created the hard
link.
Kasper
On Dec 1, 2009, at 5:22 PM, Kasper Daniel Hansen wrote
Vince
I am building a 64bit version of R using only Xcode 3.1.2 as well as
the gfortran that Simon provides. This is on OS X 10.5.7. It is an old
installation, but I have only installed very few programs from source
(and have never used fink nor darwinports), like graphviz and similar
Hi Vladimir
Based on your output my guess is that you need to install the
developer tools (compilers, make, etc.). These tools reside on your OS
X system install DVDs, but are not installed automatically.
Kasper
On Jun 21, 2009, at 19:58 , Vladimir Torres wrote:
MacBook Intel Core Duo
What do you mean by running R.app directly from the X11 window?
We would also like to get some details on what version of Rgraphviz
you are using as well as what version of Graphviz you are using and
how you installed the two versions? graphvizVersion() might be useful.
It might also be
There are Apple specific defines in GCC. As far as I know
__APPLE__
means compiling on Apple hardware, whereas
__APPLE_CC__
means compiling using an Apple supplied compiler.
However, the first one seems sometimes to be used when the last one is
intended.
Kasper
On Mar 19, 2009, at 21:50 ,
My tar is in /usr/bin. According to your output, it is not present in
that directory. As Kjell, I have Xcode install, but I am pretty sure
tar is available on a standard (non-Xcode installed) Mac system -
otherwise the binary distribution would have a bit of a problem.
My conclusion is
Good to know, thanks.
Kasper
On Mar 5, 2009, at 9:49 , Simon Urbanek wrote:
On Mar 5, 2009, at 12:47 , Simon Urbanek wrote:
On Mar 5, 2009, at 4:42 , Prof Brian Ripley wrote:
To Simon: this was Mac-specific as only on Macs does
install.packages use this piece of code (and we had a
Regarding PATH and R.app. It is a well-known issue on OS X that R.app
does not inherit the environment variables set in the shell (for
example in ~/.profile or ~/.bashrc).
For example, it is complete standard that you get different output from
# echo $PATH
and
R Sys.getenv(PATH)
The
Yes, remove .Rhistory and possibly also .RData and your .Rprofile
Kasper
On Mar 2, 2009, at 8:29 , William Revelle wrote:
Mary,
A somewhat similar problem happened to me in January.
Following the advice of Byron Ellis , I removed the .Rhistory file
At 4:45 PM -0800 1/1/09, Byron Ellis
It is considered extremely bad behavior to report something like this
as bug (see recent post on R-devel).
Kasper
On Mar 2, 2009, at 8:50 , Mary Meyer wrote:
Thank you, that worked! It's now again running happily.
Best,
Mary
On Mar 2, 2009, at 9:45 AM, Kasper Daniel Hansen wrote:
Yes
Ask on R-help, but note that this is more of a statistics question
that a R question.\
Kasper
On Jan 27, 2009, at 6:08 , Etienne Toffin wrote:
Hi,
I've made a research about how to compare two regression line slopes
(of y versus x for 2 groups, group being a factor ) using R.
I knew
/
R.framework/Versions/2.9/Resources and /Library/Frameworks/
R.framework/Resources is a symbolic link to /Library/Frameworks/
R.framework/Versions/2.8/Resources.
Patrick
Kasper Daniel Hansen wrote:
I have two version of R installed in
/Library/Frameworks
version 2.8 patched and version 2.9 (R-devel
Put the following in your ~/.Rprofile (you might have to create this
file) to revert to the old behaviour
Sys.setenv(SWEAVE_STYLEPATH_DEFAULT = TRUE)
And this question have been asked many times recently on various email
lists.
Kasper
On Nov 16, 2008, at 8:53 , Denis Chabot wrote:
The standard explanation for this is that the GUI is unresponsive
until all selected packages have been downloaded and installed - this
may take quite a while. The fact that it works for you one package at
a time corroborates this.
Kasper
On Nov 2, 2008, at 13:57 , Timothy Bates wrote:
Loren, the easiest solution to your problem is to save objects in
individual files, like
save(BigComputation, file = BigComputation.rda)
that way you just have one object in each file and can use the file
time stamp.
Using attributes are of course another approach.
Kasper
On Oct 30,
Thanks to Simon for all the work on the compilers.
I can see that the compilers and instructions have been updated on the
tools page (r.research.att.com/tools). It seems as if the
recommended way for us Leopard users is to use
gcc 4.2 (Apple Inc. build 5564) for Mac OS X 10.4 (Tiger):
, [Ricardo Rodriguez] Your XEN ICT Team wrote:
Hi Kasper
Kasper Daniel Hansen wrote:
Ok, this is clearly wrong. As you can see in your output, this PATH
does not contain /usr/bin where tar is. In my case I get
R Sys.getenv(PATH
Clearly tar is not in the PATH. Your PATH variable can be different
from application to application and in this case however you have
started R (GUI?), tar is not being picked up. And which tar are you
using? Xcode, FInk, Macports.
Kasper
On Oct 3, 2008, at 14:14 , Ben Bolker wrote:
Ok, I guess I was being a bit unhelpful. You are trying to install a
binary from Bioconductor. This should work out of the box. What
version of R are you using. If you are doing 64bit computing you need
to install from source. Of course, that does not change that tar seems
to be missing
wrote:
Hi Kasper,
Kasper Daniel Hansen wrote:
Clearly tar is not in the PATH. Your PATH variable can be different
from application to application and in this case however you have
started R (GUI?), tar is not being picked up. And which tar are you
using? Xcode, FInk, Macports.
I must
Ok, ignore this post, I found the culprit and it is me.
Kasper
On Aug 29, 2008, at 3:24 PM, Kasper Daniel Hansen wrote:
Thanks a lot, that was very helpful.
On a side note I am surprised to see that R-devel's configure script
picks up
/usr/local/bin/gfortran-4.2
instead of
/usr/bin
I want to get a suitable gfortran for Xcode 3.1 under Leopard in order
to build x86_64
Despite reading the tools and building pages on r.research.att.com
I am still confused. The comment on GCC 4.2 on the building page
seems to indicate that I should look under the Alternative section
on
Jan send out the email below quite a while ago. Now 10.5.4 has been
released as well as Xcode 3.1 and I don't see any sign of gfortran (at
least when I read the release notes, I have not installed Xcode 3.1
yet). I am not following the apple developer email lists, but could
someone please
Like Sean is aying, you most likely are using _way_ more memory than
1.2 GB.
However, if you a re running 32bit R (which is the case if you use the
CRAN binary) R can only access 2GB, so you can squeeze a little more
out of your machine by switching to a 64bit version of R. You can
check
Ok, I will take this over to the R-SIG-Mac list and cc Herve who is
the build manager of Bioconductor. I suspect that something is wrong
with the Bioconductor builds.
Synopsis: Jason is running R-2.7.0 from CRAN on an Intel macbook
running 10.5.3. He says he has Xcode 3.0 installed. He
Daniel Hansen wrote:
Ok, I will take this over to the R-SIG-Mac list and cc Herve who is
the build manager of Bioconductor. I suspect that something is
wrong with the Bioconductor builds.
Synopsis: Jason is running R-2.7.0 from CRAN on an Intel macbook
running 10.5.3. He says he has Xcode
On Jun 19, 2008, at 11:42 AM, Simon Urbanek wrote:
Are you? Did you actually test it? I'm pretty sure that you're
wrong, because the commands below compile libintl.8.dylib(!) and
thus won't help (that one is fine and part of your Gtk binary). I'm
still puzzled at where the BioC team gets
On Jun 13, 2008, at 11:24 PM, Vadim Patsalo wrote:
Hello Everyone,
I was hoping to make the following feature request: I've noticed
that R.app uses the standard Ctrl-a and Ctrl-e to jump to the
beginning and the end of the line on the console and in the editor.
However, I have not found
Simon might correct me on this.
My belief is that any 64bit build on Mac OS X has been considered
experimental for a while and probably will in the foreseeable future.
One of the problems is that if the CRAN build supports x86_64, then
all binary packages (CRAN and Bioconductor) should be
Read ?download.file - and browse the archives. This question has been
asked 100s of times (perhaps not on this list but certainly on other R-
related lists).
Kasper
On May 1, 2008, at 5:48 AM, Shelah Morita wrote:
Hi,
Does no one have a solution for me? Or does this not make sense?
Write to the package maintainer or the author to ask about this.
Kasper
On Mar 26, 2008, at 6:02 PM, Cameron Redsell-Montgomerie wrote:
I am currently writing a paper comparing various packages in
different programs, and how they handle GEE. I was wondering if you
could help me out a bit.
I recently updated to Leopard and have not gotten the error Martin is
alluding to when using x11 from the command line. However I have been
careful to remove any user-specific setting of the DISPLAY variable as
this is no longer required nor recommended under leopard.
My guess is that you
Do anyone have experience using the intel compilers and the MLK
library with R? If so, what is the speedup approximately (I know this
depends heavily on what task you are doing, but I am just looking for
some general impressions).
Kasper
___
Why are you using R-2.5, when you try to compile R-2.6.2 from source?
This seems to indicate that you do not really need an older version...
Without being sure I would think that if you use 2.6.2 on Tiger you
can use CarbonEL to get 64bit display quartz device functioning from
the command
The way to get control over ESS if you are stuck in a big loop is to do
C-g
Then R will run happily and you can continue writing in other
buffers. You will not be able to stop the process short of actually
killing R - if you do for example
C-c C-c
which sends an interrupt, R will not
I am wondering what compiler I need to use under Leopard when I
install packages from source using the experimental build of R-2.6-
branch for x64 (from Simon's site), when I am running leopard. To my
mind there are two options
1) The standard Xcode 3.0 gcc compiler
2) The preview of gcc 4.2
Let me preempt Simon a bit and suggest that you give us very precise
information about what happens. Especially these days since Mac
actually covers 4 different architectures, several versions of the OS
and several compilers.
So
1) Get a crash report and send it to Simon. If it is so easy
Hi All
I am helping organize a course where we will use Rgraphviz. One of
our future participants are using an Intel Macbook and has trouble
installing Rgraphviz. Clearly my possibilities for probing around are
rather limited since the course has not started, but this is what he
reports
- install Xcode (preferably 2.4.1 or 2.5)
Cheers,
Simon
On Nov 1, 2007, at 1:55 PM, Kasper Daniel Hansen wrote:
Hi All
I am helping organize a course where we will use Rgraphviz. One of
our future participants are using an Intel Macbook and has trouble
installing Rgraphviz. Clearly my
, at 2:45 PM, Kasper Daniel Hansen wrote:
The binaries from CRAN only works with OSX 10.4.4 and higher (as it
clearly says on the download page). You will probably need to
compile from source.
Kasper
On Sep 24, 2007, at 2:20 PM, Jeff Morris wrote:
Hello all. I am a new Mac user and am having
The binaries from CRAN only works with OSX 10.4.4 and higher (as it
clearly says on the download page). You will probably need to
compile from source.
Kasper
On Sep 24, 2007, at 2:20 PM, Jeff Morris wrote:
Hello all. I am a new Mac user and am having a problem installing
R. I
am
For Rgraphviz we need to know what version of Graphviz you are using
and how you installed it.
Kasper
On Sep 16, 2007, at 10:26 AM, William Revelle wrote:
Hi,
The packages polycor and Rgraphviz have a bus error (and R crash)
when running their examples under 2.6.0 alpha. They both work
On Jul 2, 2007, at 9:57 PM, Weiwei Shi wrote:
Hi, there:
I am re-compiling Package supclust for some small modification for my
own purpose. After running R CMD check, I got the following
messenge:
nicebaby:~/Documents/projects/R_customerized_packages wshi$ R CMD
check supclust
*
Hi
I am using the CRAN binary of R-2.5.0 together with the supplied
gfortran and the latest Xcode on a PPC mac running OS X 10.4.9. In my
~/.profile I have
export R_ARCH=/ppc
It seems as if the default of R nowadays is to build universal
binaries of packages. At least when I try to
Also, if the aim is just to provide an R package for collaborators or
such, you can use
http://129.217.207.166/index.htm
Read the disclaimer though, in case your package contains sensitive
data.
Kasper
On May 12, 2007, at 7:01 AM, Simon Urbanek wrote:
On May 12, 2007, at 8:24 AM, Klaus
On May 12, 2007, at 12:43 PM, cstrato wrote:
Dear all
In my package I need to get the path to the shared library
mypackage.so.
Eh, why do you need to do so? I am not saying there could not be use
cases, but most standard uses should not need to do this.
Kasper
Currently, I use the
On Apr 24, 2007, at 9:49 AM, James wrote:
On Apr 24, 2007, at 10:29 AM, Rob J Goedman wrote:
Hi James,
Could you try if it retains other preference setting changes, e.g.
for dragdrop or save on exit?
It doesn't appear to retain these settings either.
These settings are
R-2.5.0 is soon to be released. From comments on this mailing list it
seems that the default compiler will move from a custom install of
GCC-4.0.3/gfortran to the version of GCC 4.0.1 supplied with Xcode
2.4.1 together with a gfortran compiler from say hpc.sourceforge.
In preparation for
On Apr 5, 2007, at 8:56 PM, Simon Urbanek wrote:
On Apr 5, 2007, at 9:07 PM, Kasper Daniel Hansen wrote:
R-2.5.0 is soon to be released. From comments on this mailing list
it seems that the default compiler will move from a custom install
of GCC-4.0.3/gfortran to the version of GCC
Your example works fine for me on Mac OS X ppc 10.4.8 with R-2.4.1
and graphviz 2.12 compiled from source.
You can check your graphviz version from R by doing
R graphvizVersion()
You have an error in the Graphviz version reported below, I assume
you are talking about Rgraphviz' version. When
PROTECTED] writes:
On Sun, 17 Dec 2006, Kasper Daniel Hansen wrote:
For package development I would like to compile
with the -Wall flag. What is the best way to add
this?
Probably the easiest [courtesy of Kurt]:
mkdir ~/.R
echo CFLAGS=-g -O2 -Wall -pedantic ~/.R/Makevars
Should I just edit
Hi
I am using the CRAN binary of R.
For package development I would like to compile with the -Wall flag.
What is the best way to add this? Should I just edit
R.Framework/Resources/ppc/etc/Makeconf
(I am on a G4), and add it to the CFLAGS, CXXFLAGS, FFLAGS variables.
Or is there a better
On Dec 17, 2006, at 12:35 PM, Paul Roebuck wrote:
On Sun, 17 Dec 2006, Kasper Daniel Hansen wrote:
For package development I would like to compile
with the -Wall flag. What is the best way to add
this? Should I just edit
R.Framework/Resources/ppc/etc/Makeconf
(I am on a G4), and add
On Dec 17, 2006, at 2:55 AM, Emmanuel Sharef wrote:
On Dec 16, 2006, at 6:50 PM, [EMAIL PROTECTED] wrote:
I get the following error when trying to compile some Fortran code
to a shared object using R CMD SHLIB:
R CMD SHLIB test3.f
gfortran-4.0 -arch ppc -fPIC -g -O2 -c test3.f -o
I will revive this old thread due to me having recently spend
considerable time in the internals of (R)Graphviz. So someone besides
me might benefit from all that time :)
Essentially the conclusions are
1) There is a bug in Graphviz 2.4, 2.6 and 2.8 that makes it
impossible to layout a
I recently upgraded to R-2.4.0 and it seems that something broke the
older framework I had installed (2.3.1 - or should I just say 2.3).
Specifically it seems that some of the symlinks in /Library/
Frameworks/R.Framwork/Versions/2.3 was missing. I then (re)installed
2.3.1 from the -mini.dmg
I have now played around a little with Rswitch, and I think it is a
great idea. It makes it very easy to test packages in the devel
version of R (which I am doing right now :).
It seems to be very close to be extremely useable for us package
developers, but I have some questions/suggestions
Hmm, in case you installed Bioconductor from source, did you remember
to install the version of GCC that ships with the CRAN version of R?
That, or some other installation mishap is all I can think of.
Kasper
On Aug 10, 2006, at 6:33 PM, Olga Vitek wrote:
Dear All,
I am having difficulty
On May 24, 2006, at 1:21 PM, Paul Roebuck wrote:
On Wed, 24 May 2006 Herve Pages wrote:
Quoting Simon Urbanek [EMAIL PROTECTED]:
On May 23, 2006, at 10:31 PM, Herve Pages wrote:
[SNIP]
but they should be (in this order):
http://bioconductor.org/packages/bioc
Having said that, affxparser does not work on Mac OS using the
version of gcc shipped with R-2.3.0. We are working on it.
/Kasper
On May 12, 2006, at 8:39 AM, Simon Urbanek wrote:
On May 12, 2006, at 8:36 AM, Booman, M wrote:
Hi Sean,
Thanks for your reply. I managed to install most of
If you are installing from the binary obtained from CRAN, choose
customize during installation. That will give you the option of
installing the tcltk stuff. And while you are at it, grab the g77 as
well. And you probably also want to use gcc 3.3, by giving the command
sudo gcc_select 3.3
You need to get the exact right version of graphviz before installing
Rgraphviz. Around December I spent most of a day getting it to work.
At that time the current stable version was 2.6 and the devel version
was 2.7. I had to use the devel version, but then it worked out of
the box so to
As several people have replied, Emacs is by far the preferred choice
for most serious R developers. Using the ESS package with Emacs
provides you with an interface to R which goes beyond just syntax
highlightning. An example (which Simon has already described): you
can press a key
Well, I guess a first thing would be to place them at the CRAN site
under other where WinEdt and ESS gets linked to, and the a2ps
stylesheets are.
Having something in the MacOSX FAQ world make sense as well.
Kasper
On Dec 6, 2005, at 1:12 PM, stefano iacus wrote:
It would be nice to
Hi
I have source compile my own version of R on MacOS Tiger.
I have a problem with path.expand and dirname. On Linux and the
binary version of R
path.expand(~/foo/)
exapnds to something like
/Users/kdh/foo
On my homebuild I just get
~/foo
I do not have gettext installed.
I assume
On Aug 30, 2005, at 8:21 PM, Seth Falcon wrote:
Hi Kasper,
On 30 Aug 2005, [EMAIL PROTECTED] wrote:
Hi
I have source compile my own version of R on MacOS Tiger.
I've also compiled R from source on Tiger. I get:
path.expand(~seth)
[1] /Users/seth/~seth
path.expand(~/tmp)
[1]
81 matches
Mail list logo