I recently upgraded to F24, and now in my R session I cannot get a few packages to load, sp. reshape2, latex2exp, knitr, and others.
The initial problem I found was that F24 uses libicu56 whereas these packages expect libicu54. I followed a suggestion in this thread to set the symbolic links with the desired version:
ln -s /usr/lib64/libicui18n.so.56 /usr/lib64/libicui18n.so.54
ln -s /usr/lib64/libicuuc.so.56 /usr/lib64/libicuuc.so.54
ln -s /usr/lib64/libicudata.so.56 /usr/lib64/libicudata.so.54
That initial error went away, but now I have this:
Error in dyn.load(file, DLLpath = DLLpath, ...) :
unable to load shared object '/home/uname/R/x86_64-redhat-linux-gnu-library/3.3/stringi/libs/stringi.so':
/home/uname/R/x86_64-redhat-linux-gnu-library/3.3/stringi/libs/stringi.so: undefined symbol: _ZTIN6icu_548ByteSinkE
This leads me to the stringi package for R, but I cannot get it to load - it gives the same error.
I have updated F24 and all the R packages as well.
Any ideas?
I was able to install the package stringi on fedora 24 by downloading the tar.gz package from CRAN and then run the following command:
R CMD INSTALL stringi_1.1.1.tar.gz --configure-args='--disable-pkg-config'
That just happened to me following an update of icu (Gentoo). Another solution is to remove and install again stringi, as it is looking for a specific library file that does not exist anymore.
remove.packages('stringi')
install.packages('stringi')
If your .Rprofile triggers library(stringi), then you must start a session using R --vanilla in order to do this, else it will keep failing.
The other solution with --disable-pkg-config works as well. It will make stringi build icu for itself rather than rely on the system's (the source ships with a copy of icu).
Related
I am working on Mac OSX, using v3.6.3 of R and using renv. In Rstudio and R, I can load the libraries of my installed packages, e.g library(ggplot2) works. However when I run a script using Rscript I get the message
Error in library(ggplot2) : there is no package called ‘ggplot2’
According to this SO answer, I need to make sure that the value of
Sys.getenv('R_LIBS_USER') in R.exe
is the same as the value of
Rscript.exe -e ".libPaths()"
But the value is the same, both are pointing to the renv-system-library in my project folder.
So how do I fix this?
It may be better to specify the lib.loc in library call
library(ggplot2, lib.loc = '/path/where/library/is/located')
I managed to solve this. akruns answer was useful, it did not work, but pointed me incorrect direction. The answer did not work because using it, I received the following error:
Error: package or namespace load failed for ‘ggplot2’:
.onLoad failed in loadNamespace() for 'pillar', details:
call: utils::packageVersion("vctrs")
error: there is no package called ‘vctrs’
Now vctrs was in the '/path/where/library/is/located' so I think dependent packages were not being loaded from that path but the default for Rscript. Putting a print(.libPaths() in the script gave
"/usr/local/Cellar/r/3.6.3_1/lib/R/library"
instead of
[1] "/Users/Chris/Sites/app_name/renv/library/R-3.6/x86_64-apple-darwin18.7.0"
[2] "/private/var/folders/5_/p_yl0439059b7_jdqzrm0hr40000gr/T/RtmptdHcWN/renv-system-library"
for .libPaths() in Rstudio. Looking at the ruby program that was actually running the Rscript program, I found it was being run with the --vanilla option, i.e
Rscript --vanilla script_name
Removing the --vanilla option fixed the problem. I think that scripts with the --vanilla option stopped working because I reinsalled R using brew to fix another problem I was having and as part of this issued this command:
brew link --overwrite r
I am trying to install the forecast package which depends on uroot, which apparently have been written to require a GPU?
install.packages("uroot") yields the following error. Has anyone found this issue and may suggest a work around? I am using ubuntu 16.04.
I have the file in question located here: /usr/local/cuda-8.0/ and I added to my path export PATH=/usr/local/cuda-8.0/:$PATH
Error in dyn.load(file, DLLpath = DLLpath, ...) :
unable to load shared object '/home/rstudio2/R/x86_64-pc-linux-gnu-library/3.2/uroot/libs/uroot.so':
libcudart.so.8.0: cannot open shared object file: No such file or directory
Error: loading failed
Execution halted
ERROR: loading failed
* removing ‘/home/rstudio2/R/x86_64-pc-linux-gnu-library/3.2/uroot’
Warning in install.packages :
installation of package ‘uroot’ had non-zero exit status
Just in case this is useful for someone, I stumbled with the same problem trying to install "imputeTS" with R version 3.5.1 on Fedora 27 with CUDA 10.0 correctly installed an setted in PATH.
for me it just happened that I was running install.package("imputeTS") (which depends on forecast) in a R console as root, installing the packages to the user home solved the problem. It's seems that the libraries are searched in different places according to the user running the script, it's important to note that.
I just migrated to Linux. On an Ubuntu 18.04 setup, using install.packages with dependencies set to TRUE succeeded with no complaints. (No GPU on mu box.)
install.packages("forecast", dep=TRUE)
# there are quite a few dependencies and you happened to be missing one
packageDescription()
#-----------includes this line----------
Imports: colorspace, fracdiff, ggplot2 (>= 2.2.1), graphics, lmtest,
magrittr, nnet, parallel, Rcpp (>= 0.11.0), stats,
timeDate, tseries, urca, uroot, zoo
You can try to modify file "/etc/rstudio/rserver.conf", add "LD_LIBRARY_PATH" env, like:
rsession-ld-library-path=/usr/local/cuda/lib64
Actually, When you open session of "RStudio Server" from browser, the session does't get env from the ".bashrc".
I recently "upgraded" from OSX Mountain Lion to Yosemite and from R 3.1.3 to 3.2. Immediately after the upgrade, when I opened R or RStudio I got a pop-up message saying that I needed to install Java 6. In addition, loading rJava or any package that depends on rJava (e.g., xlsx) caused RStudio to crash (R also crashed when I tried this by opening R.app directly).
After trying a few fixes found on Stack Overflow and elsewhere (more details below), I am at a point where loading rJava or any package that depends on rJava no longer causes R to crash, but results in the following error:
library(rJava)
Error : .onLoad failed in loadNamespace() for 'rJava', details:
call: dyn.load(file, DLLpath = DLLpath, ...)
error: unable to load shared object '/Library/Frameworks/R.framework/Versions/3.2/Resources/library/rJava/libs/rJava.so':
dlopen(/Library/Frameworks/R.framework/Versions/3.2/Resources/library/rJava/libs/rJava.so, 6): Library not loaded: #rpath/libjvm.dylib
Referenced from: /Library/Frameworks/R.framework/Versions/3.2/Resources/library/rJava/libs/rJava.so
Reason: image not found
Error: package or namespace load failed for ‘rJava’
However, if I invoke R from the command line and then load rJava or any package that depends on rJava, it seems to work (or at least I don't get any error messages).
I've tried a number of different attempted fixes, some of them a few times, and can't quite remember exactly what I did in what order (didn't realize this would be such a morass and wasn't really keeping track), but here's the gist of it:
Added the following to my .bash_profile (per this SO answer):
export JAVA_HOME="/usr/libexec/java_home -v 1.8"
export LD_LIBRARY_PATH=$JAVA_HOME/jre/lib/server
Reconfigured java from the command line as follows:
sudo R CMD javareconf -n
Checked options("java.home") and discovered this was set to NULL. I tried setting it to the following (per this SO question):
options("java.home"="/Library/Java/JavaVirtualMachines/jdk1.8.0_45.jdk/Contents/Home/jre")
Installed the latest Java Development Kit and reinstalled rJava from source (can't remember where I found that one).
At some point while trying all of these, I was able to load rJava without crashing R, but instead got the error message posted above. In addition, when I quit RStudio, it would seem to close normally, but then an "RStudio quit unexpectedly" message would pop up, indicating that the program had crashed while trying to close.
I finally decided to install Java for OS X 2014-001 (Java 6), as I seemed to be running out of options. Now, when I opened R or RStudio the "This software needs Java 6" pop-up message no longer appeared. However, I was still getting the .onLoad failed in loadNamespace() for 'rJava' error message posted above.
In reviewing some of the posts I'd already looked at, I noticed another SO answer that I'd missed before, which recommended opening RStudio with the following command line code that gives RStudio the correct path to java:
LD_LIBRARY_PATH=$(/usr/libexec/java_home)/jre/lib/server: open -a RStudio
That opened an RStudio window and I was also able to load rJava and packages that depend on it without getting an error.
Lastly, I tried running R from the command line (which I hadn't done before). It turns out that on the command line, loading rJava or any package that depends on rJava works and does not throw any errors.
So, I can now get rJava to work if I open RStudio from the command line with the code that gives RStudio the java path (as noted above). However, I'd like to find a way to fix the underlying problem, whatever it may be, so that RStudio can be opened in the usual Mac way, without needing a command line kludge. I'm also concerned that having an old version of Java installed could cause problems down the road.
Does anyone have any ideas about how to diagnose and solve this issue?
I had the same problem and went through the same steps as you. The final step to allow starting RStudio through Finder/Spotlight was to link libjvm.dylib to /usr/local/lib:
sudo ln -f -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
-f flag is added to force overwriting existing file/link
On OSX El Capitan 10.11, the user doesn't have permission to write to /usr/lib. So instead, use the following command:
sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
You can load the libjvm.dylib explictly, by run like
dyn.load('/Library/Java/JavaVirtualMachines/jdk1.8.0_66.jdk/Contents/Home/jre/lib/server/libjvm.dylib')
library(rJava)
On macOS High Sierra (10.13.1) and Java Version 9 you have to use a slightly different JVM path (notice the missing jre folder in the path compared to the instructions for earlier Java versions):
sudo ln -f -s $(/usr/libexec/java_home)/lib/server/libjvm.dylib /usr/local/lib
You also have to notify R about the JVM:
MY_R_VERSION=$(Rscript -e "cat(with(R.version, sprintf('%s.%s', major, substring(minor, 1, 1))))")
ln -s /usr/local/lib/libjvm.dylib /Library/Frameworks/R.framework/Versions/$MY_R_VERSION/Resources/lib/
Following command works:
sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
After that, in RStudio, loading rJava works through loading of the "xlsx" package.
library("xlsx")
P.S. #1 Environment: Mac OS X El Capitan 10.11.3+ with RStudio 0.99.491+ and R 3.2.3+. (I've now tested this also on macOS Sierra (10.12) and R.3.3.1.)
P.S. #2 I find that openxlsx is much faster and doesn't rely on Java to work, so I'm now using that package. Hope that helps others.
You should be able to use the CRAN rJava, without needing to recompile rJava or do any additional steps by linking the shared library to the R frameworks library directory.
sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /Library/Frameworks/R.framework/Libraries
This is how I configured it working properly on two Macs running Mac OS X El Capitan (10.11.6):
Uninstall 'rJava' by issuing the following commands in a terminal window:
Rscript -e 'remove.packages("rJava")'
sudo Rscript -e 'remove.packages("rJava")'
Download and install the Java software from Oracle: https://www.java.com/en/download/mac_download.jsp
Add the following lines to /Users/<userid>/.bashrc using your favorite editor:
# Set JAVA_HOME so rJava package can find it
export JAVA_HOME=$(/usr/libexec/java_home -v 1.8)/jre
Close and restart any terminal, R, and RStudio windows (to pick up the changes to .bashrc).
Run the following command in a Terminal window:
sudo ln -sf $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
Run the following command in a Terminal window:
sudo Rscript -e 'install.packages("rJava", repos="http://rforge.net", type="source")'
Here is the easy steps for it:
remove the rJava package: remove.packages(rJava)
close R
install latest Java on you mac
open terminal and type this command: sudo R CMD javareconf
Open R and install rJava with this command:
install.packages("rJava", dependencies=TRUE, type="source")
The image not found issue exists with a brand-new install on OSX High Sierra with the latest Java SE 10 JDK.
I was able to solve the path issue with rJava using the fix found on the rJava Github issues page: https://github.com/s-u/rJava/issues/78
R CMD javareconf JAVA_CPPFLAGS=-I/System/Library/Frameworks/JavaVM.framework/Headers
Here what worked for me on MAC:
in your ~/.profile or ~/.bashrc add this line:
%export JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_102.jdk/Contents/Home
1.1 % source .profile (or % source .bashrc)
% sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
remove.packages(rJava)
remove.packages(Rweka)
From the terminal enter this command:
%sudo R CMD javareconf
install.packages("rJava", dependencies = TRUE, type = "source")
install.packages("rJava", dependencies = TRUE, type = "source")
Load your library("rJava", "RWeka")
Best of luck.
Maybe another simple answer which does not touch your filesystem:
$ install_name_tool -add_rpath /Library/Java/JavaVirtualMachines/jdk1.8.0_131.jdk/Contents/Home/jre/lib/server /Library/Frameworks/R.framework/Versions/3.4/Resources/library/rJava/libs/rJava.so
(Substitute jdk1.8.0_131.jdk with your JDK path.)
Check the location of libjvm.dylib file.
Try this, in my case this worked:
dyn.load('/Library/Java/JavaVirtualMachines/jdk-10.jdk/Contents/Home/lib/server/libjvm.dylib')
library(rJava)
I went through all the troubleshooting I could find, then installed jdk-11.0.1_osx-x64_bin.dmg from: Oracle downloads
Everything worked perfectly after that.
Check the version of the library that was not loaded when you run library('rJava') and match it to the java version you need installed.
I noticed that 'rJava' is pretty strict about JDK version while loading in Rstudio. In my case (Mac OS. 10.14.6 - Mojave) I had installed jdk-13 which was tested against jdk-11.
I have linked my JDK version to a non-existent directory that was assumed to be valid by Rstudio (JDK 11):
sudo ln -sf /Library/Java/JavaVirtualMachines/jdk-13.jdk/ /Library/Java/JavaVirtualMachines/jdk-11.0.1.jdk
and suprisingly it works like a charm...
BTW., there was no problem to run 'rJava' following the previous instructions in native R.
Here you can download the legacy Java version 6 for El Capitan: https://support.apple.com/kb/DL1572?viewlocale=en_US&locale=en_US
After doing so, the rJava package works for me.
This line below has solved the same problem I was having with rJava package as some others in this discussion did. I'm sure there are more than one solution to this problem and I sincerely thank them all for their contributions because sometimes one line like the one below saves a lot of time from going to trash!
sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
Run the following on the command line: sudo R CMD javareconf
Several solutions above have mentioned this, but they also suggest that one remove and subsequently re-install the rJava package. I found those additional steps to be unnecessary.
You can change the reference in R by using the following command:
sudo install_name_tool -change /Library/Java/JavaVirtualMachines/jdk-11.0.1.jdk/Contents/Home/lib/server/libjvm.dylib /Library/Java/JavaVirtualMachines/jdk-13.0.1.jdk/Contents/Home/lib/server/libjvm.dylib /Library/Frameworks/R.framework/Versions/3.6/Resources/library/rJava/libs/rJava.so
I have a general question about how to effectively load any kind of external package into R. I have found that many sources detailing this information are simply insufficient, but I will leave out mentioning those respective URLs. I have successfully loaded external packages into R before, but I am having difficulties loading the fgui package.
In my specific case, I cannot load the library fgui into R Studio. First I download directly from CRAN with the command:
install.packages("fgui", lib="~/Documents/R_dir")
This works fine for me. My working directory is set to ~/Documents/R_dir
The next command I use is:
library("fgui", lib.loc="~/Documents/R_dir")
The error I get is:
Error : .onLoad failed in loadNamespace() for 'tcltk', details:
call: dyn.load(file, DLLpath = DLLpath, ...)
error: unable to load shared object '/Library/Frameworks/R.framework/Versions/3.1/Resources/library/tcltk/libs/tcltk.so':
dlopen(/Library/Frameworks/R.framework/Versions/3.1/Resources/library/tcltk/libs/tcltk.so, 10): Library not loaded: /opt/X11/lib/libX11.6.dylib
Referenced from: /Library/Frameworks/R.framework/Versions/3.1/Resources/library/tcltk/libs/tcltk.so
Reason: image not found
Error: package or namespace load failed for ‘fgui’
I have examined the thread "Error when trying to load library(ggplot2) in R" but it did not answer my question.
Please help me with any suggestions to get the package fgui to work properly in R Studio and feel free to address the more general issue of effectively loading external packages into R.
tcltk is not a typical package. It's probably already installed on your system and it cannot be installed using the usual install.packages procedure because it is not on CRAN (just like the base packages that come with R, like stats, etc.). Also, tcltk2 is not the same as tcltk; that is a package that expands upon (and depends on) the functionality of tcltk.
So, start with this:
> capabilities("tcltk")
tcltk
TRUE
If that comes back FALSE it means that your R was not built with tcltk support. You don't say what OS you're using, so it's a little bit difficult to help you.
That said, if you're on Linux, you need to rebuild R with tcltk support. This should be relatively straightforward. Basically, make sure Tcl and tk are available on your system in a standard location (something like sudo apt-get install tcl8.5-dev tk8.5-dev) and then rebuild R per usual procedures.
If you're on Mac OS, you're encountering the same error that John Fox describes here in relation to Rcmdr. In short, it's probably one of two things:
A file permission issue related to the directory where you have Tcl installed. To quote Fox: "You can verify the source of this problem by issuing the following command at the R command prompt:
system("ls -ld /usr/local /usr/local/lib /usr/local/lib/libtcl*")
# ls: /usr/local/lib: Permission denied
# ls: /usr/local/lib/libtcl*: Permission denied
# drwx------ 8 root wheel 272 Sep 24 10:21 /usr/local
If the output of the system command looks like the above, you should go to Terminal and do the following: sudo chmod -R a+rX /usr/local and enter your password to change directory permissions.
You have an out-dated version of X-Windows. Try running Mac OS Software Update.
It may also be solvable by creating a symbolic link as described in this answer.
This should never come up on Windows, unless you built R from source without tcl/tk support.
On Arch Linux, tk is a separate package from tcl and both need to be installed:
sudo pacman -S tcl tk
This is given that the output from capabilities("tcltk") with R is TRUE (which means your R is compiled with tcltk support as noted previously in the answers here, this should be the case if you installed it from the arch repos).
The official installation instructions for R say:
Note: the use of X11 (including tcltk) requires XQuartz to be
installed since it is no longer part of OS X. Always re-install
XQuartz when upgrading your OS X to a new major version.
So: Simply (re)install XQuartz if you want to use the tcltk package in R.
delete the X11 file on location : /opt
delete the X11 on application unitiltiy
reinstall X11 again
Problem solved...
For Mac-OS users:
I struggled with this problem recently and found that R-3.1.2 has a bug, which assumes that otool is installed when it is not.
R-3.0.0 (and later) ships with Tck/Tk 8.6.0, so you shouldn't need to install that.
Simply upgrading to a more recent version of R (3.4.2) fixed the problem for me.
I had the same issue and went down a rabbit hole before discovering a simple fix. If you install R from bianry it will include tcltk by default. Just run the installer with the default installation. You can get the .dmg here: https://cran.r-project.org/bin/macosx/
Easiest answer for Mac users:
1- Install tcl-tk using brew:
brew install tcl-tk
2- Run the following code:
ln -s /usr/local/Cellar/tcl-tk/$version/lib/libtcl8.6.dylib /usr/local/lib/libtcl8.6.dylib
ln -s /usr/local/Cellar/tcl-tk/$version/lib/libtk8.6.dylib /usr/local/lib/libtk8.6.dylib
Problem solved!
I'm running from Ubuntu 16, using R version 3.4.1. I have dplyr installed and can load it when either I am running from RStudio or when I sudo into R from the terminal. However, if I run R without root permission, I cannot load dplyr due to the following error:
Error: package or namespace load failed for ‘dplyr’ in dyn.load(file, DLLpath = DLLpath, ...):
unable to load shared object '<user-directory>/R/x86_64-pc-linux-gnu-library/3.4/Rcpp/libs/Rcpp.so':
<user-directory>/anaconda3/lib/R/bin/exec/../../lib/../.././libstdc++.so.6: version `GLIBCXX_3.4.20' not found (required by <user-directory>/R/x86_64-pc-linux-gnu-library/3.4/Rcpp/libs/Rcpp.so)
(where above I've replaced the /home/MY_USER_NAME dir with <user-directory> for some mild anonymization). I've seen a couple posts that may be tangentially related, but none of those remedies worked for me.
I'm curious if the fix is something as simple as changing permissions or ownership of certain files, or if the issue is that the search for GLIBCXX_3.4.20 is directed into my anaconda3 directory (when not root?!)?
Thank you in advance for any insights!
Edit: This may be an issue from a more base level with my installation of R. Often I install packages using install.packages(...) within RStudio, but on occasion this results in an error that is overcome by installing with install.packages(...) after entering R as root. This was the case with both dplyr and Rcpp. I.e., the versions which I have just reinstalled were done so as
> sudo R
> install.packages( 'Rcpp', dependencies = T )
> install.packages( 'dplyr', dependencies = T )
Please make your question useful and understandable by stating how you installed these packages.
They come both from Ubuntu ie sudo apt-get install r-cran-dplyr will get a working version, and from CRAN ie install.r dplyr will too (here using my install.r from littler, use other wrappers to install.packages() at your leisure).
Sometimes your underlying libraries change so my first bet would be to simply reinstall Rcpp and then dplyr.
Your reference to /home suggests you installed from source, but as root. That is a bad idea. There a number of other questions answered here on the site about when/where/how apt-get may (or may not) be preferable so I don't need to repeat this.
Edit: In response to your edit:
Never ever do sudo R. You want to write to /usr/local/lib/R/site-library and for that you can 1.a) add yourself to the group owning that directory or 1.b) make it owned by your group if it is just you on that machine.
You could just do install.packages(c("Rcpp", "dplyr")) but you don't even need that as dplyr depends on Rcpp.
Why not use the prebuild package? Remove what you have (ie sudo R CMD REMOVE Rcpp and ditto for dplyr, then do sudo apt-get install r-cran-dplyr