Can't install R packages for Jupyter : Windows

[Diego-MX]

I have several issues installing R packages for Jupyter.

• The base command gives me this

In [1]: install.packages("readxl")
Warning message:
unable to access index for repository https://cran.itam.mx/bin/windows/contrib/3.1Warning 
message: package 'readxl' is not available (as a binary package for R version 3.1.3)

When I browse to that address, I do find the package is available.

• So I download the file and install it manually. Works on downloading but not on loading.

In [2]: install.packages("~/Downloads/readxl_0.1.1.zip", repos=NULL)
In [3]: library(readxl)
Error: package or namespace load failed for 'readxl'

• I found this post in Stackoverflow referencing the issue, so I download Rcpp and follow their solution (will not work 4 me)

In [4]: remove.packages('readxl')
In [5]: update.packages("~/Downloads/Rcpp_0.12.4.zip", repos=NULL)
In [6]: install.packages("~/Downloads/readxl_0.1.1.zip", repos=NULL)
Error: package or namespace load failed for 'readxl'

• Same error. So I decide to try to go through another installation method. Download devtools and dependencies to install them with install_github,

In [7]:  remove.packages('readxl')
In [8]:  install.packages(paste0("~/Downloads/", c("devtools_1.10.0.zip", "curl_0.9.6.zip")))
In [9]:  library(devtools)
In [10]: install_github('hadley/readxl')
WARNING: Rtools 3.3 found on the path at C:/RBUILD~1/3.3 is not compatible with R 3.1.3.

Please download and install Rtools 3.2 from http://cran.r-project.org/bin/windows/Rtools/, 
remove the incompatible version from your PATH.
Error: Could not find build tools necessary to build readxl

It seems like there's a bigger problem now with Rtools.
So it seems that I just shouldn't use readxl and go with another tool like openxlsx.
The reason why I think this is related to Anaconda is because I have another installation of R running locally on my computer, and it does install the packages as its supposed to, and I'm able to use readxl there.
If someone else has similar issues, maybe it will be worth it to try to solve it.
Cheers from Mexico.

[mingwandroid]

Conda's R does not use Rtools, instead opting to use the MSYS2 mingw-w64 toolchain (they're both mingw-w64 toolchains though and pretty similar).

We now provide prebuilt readxl, so the easiest thing to do is just:

conda install -c r r-readxl

(and if you are using conda >= 4.3 then you don't even need -c r anymore).

Otherwise, (and for other packages) to setup a dev-environment that works with R and devtools, please do the following from a cmd.exe shell in which your conda R environment is activated (or your root env if that's where you installed R/r-essentials):

conda install m2w64-toolchain m2-make r-devtools r-essentials

Then in R:

library(devtools)
options(download.file.method = "wget")
install_github('hadley/readxl')

[mikofski]

Sorry to comment on an old issue, but for anyone who lands here as I did, there are a few more steps.

• you may need to install a few more of the msys2 & mingw64 conda packages like sed, zip, gzip, tar, texinfo, diffutils, and coreutils . See R install & admin docs Appendix D on Windows Toolset: command line tools.

    $ conda install m2-sed m2-zip m2-gzip m2-tar m2-texinfo m2-coreutils m2-diffutils
  

• you may need to create a /tmp directory to %LOCALAPPDATA%\Continuum\<anaconda | miniconda>\envs\<R-env>\Library\

• you may need rsync according to R docs, but it isn't yet included in msys2 channel

There's got to be a wiki or document somewhere that explains this, but I can't seem to find it. I bet a lot of people are installing Rtools and setting BINPREF not realizing that this is unnecessary. The msys2 m2 and mingw-w64 m2w64 conda channels are a big win for Anaconda!

[mingwandroid]

If you use conda skeleton cran (which is part of conda-build) then it will attempt to generate recipes that list their dependencies. It usually works pretty well.

rsync is definitely something I wish I had added when I created the m2* packages.

I could make a metapackage called r-rtools that installs all these packages, what do you think?

I am also wondering whether to change install.packages() so that it looks for an existing conda package and failing that, uses conda skeleton cran and then conda-build and only uses the current codepath if both of these fail. It might be controversial but I think it would better serve our users than the current situation.

[mikofski]

I think the r-rtools idea is good, "rtools" was intuitively the first keyword I searched for on anaconda.org. Also having m2-rsync would be useful for other packages like git-fat that use it. Thanks for your response!

[zaphot]

Hi,

i still have problems installing (non-anaconda) R packages which require compilation.
I have installed the following anaconda R packages:

conda install -c r  r-essentials r-devtools rstudio
conda install m2w64-toolchain m2-make m2-sed m2-zip m2-gzip m2-tar m2-texinfo m2-coreutils m2-diffutils

The example from above

library(devtools)
options(download.file.method = "wget")
install_github('hadley/readxl')

only yields

Downloading GitHub repo hadley/readxl@master
from URL https://api.github.com/repos/hadley/readxl/zipball/master
WARNING: Rtools is required to build R packages, but is not currently installed.

Please download and install Rtools 3.4 from http://cran.r-project.org/bin/windows/Rtools/.
Installation failed: Could not find build tools necessary to build readxl

pulling the source code and building from command line also fails:

git clone https://github.com/tidyverse/readxl.git
R CMD build readxl
* checking for file 'readxl/DESCRIPTION' ... OK
* preparing 'readxl':
* checking DESCRIPTION meta-information ... OK
* cleaning src
* installing the package to build vignettes
Warnung: Ausführung von Kommando '"C:/Users/thorsten/Anaconda3/lib/R/bin/x64/Rcmd.exe" INSTALL -l "C:\Users\thorsten\AppData\Local\Temp\RtmpkzFvtE\Rinst2ddc3dc2218b" --no-multiarch "C:/Users/thorsten/AppData/Local/Temp/RtmpkzFvtE/Rbuild2ddc3bdf5228/readxl"' ergab Status 1
      -----------------------------------
* installing *source* package 'readxl' ...
** libs
rm -Rf readxl.dll RcppExports.o XlsWorkBook.o XlsWorkSheet.o XlsxWorkBook.o XlsxWorkSheet.o endian.o ole.o xls.o xlstool.o zip.o
c:/Rtools/mingw_64/bin/g++ -m64 -I"C:/Users/thorsten/ANACON~1/lib/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"C:/Users/thorsten/Anaconda3/Lib/R/library/Rcpp/include"   -I"C:/swarm/workspace/External-R-3.4.3/vendor/extsoft/include"     -O2 -Wall  -mtune=core2 -c RcppExports.cpp -o RcppExports.o
bash.exe: warning: could not find /tmp, please create!
/bin/sh: c:/Rtools/mingw_64/bin/g++: No such file or directory
C:/Users/thorsten/ANACON~1/lib/R/etc/x64/Makeconf:215: recipe for target 'RcppExports.o' failed
make: *** [RcppExports.o] Error 127
Warnung: Ausführung von Kommando 'make -f "Makevars.win" -f "C:/Users/thorsten/ANACON~1/lib/R/etc/x64/Makeconf" -f "C:/Users/thorsten/ANACON~1/lib/R/share/make/winshlib.mk" SHLIB_LDFLAGS='$(SHLIB_CXXLDFLAGS)' SHLIB_LD='$(SHLIB_CXXLD)' SHLIB="readxl.dll" WIN=64 TCLBIN=64 OBJECTS="RcppExports.o XlsWorkBook.o XlsWorkSheet.o XlsxWorkBook.o XlsxWorkSheet.o endian.o ole.o xls.o xlstool.o zip.o"' ergab Status 2
ERROR: compilation failed for package 'readxl'
* removing 'C:/Users/thorsten/AppData/Local/Temp/RtmpkzFvtE/Rinst2ddc3dc2218b/readxl'
In R CMD INSTALL
      -----------------------------------
ERROR: package installation failed

So do i need to install Rtools from cran or is there an equivalent available in conda? And where exactly do i have to create the /tmp folder for base environment? The path from above didn't seem to work either

[csoja]

@zaphot please open a new issue instead of commenting on an issue that has already been closed.

[mikofski]

@zaphot, I added some wikis with help on installing R packages from source or CRAN. Note I didn't use devtools or try to install from a clone, but if I grab the link from the CRAN readxl webpage I can install it from R:

(py36r) Downloads>curl -Ok https://cran.r-project.org/src/contrib/readxl_1.0.0.tar.gz
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 1890k  100 1890k    0     0   547k      0  0:00:03  0:00:03 --:--:--  550k

(py36r) Downloads>R

R version 3.4.2 (2017-09-28) -- "Short Summer"
Copyright (C) 2017 The R Foundation for Statistical Computing
Platform: x86_64-w64-mingw32/x64 (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

  Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> install.packages("readxl_1.0.0.tar.gz", repos=NULL, type="source")
* installing *source* package 'readxl' ...
** package 'readxl' successfully unpacked and MD5 sums checked
** libs
rm -Rf readxl.dll RcppExports.o XlsWorkBook.o XlsWorkSheet.o XlsxWorkBook.o XlsxWorkSheet.o endian.o init.o ole.o xls.o xlstool.o zip.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c RcppExports.cpp -o RcppExports.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c XlsWorkBook.cpp -o XlsWorkBook.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c XlsWorkSheet.cpp -o XlsWorkSheet.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c XlsxWorkBook.cpp -o XlsxWorkBook.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c XlsxWorkSheet.cpp -o XlsxWorkSheet.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/gcc  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -std=gnu99 -march=x86-64 -mtune=generic -c endian.c -o endian.o
endian.c: In function 'xlsConvertFormula':
endian.c:231:4: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
    W_ENDIAN(*(WORD *)&f->resdata[1]);
    ^
endian.c:231:4: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/gcc  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -std=gnu99 -march=x86-64 -mtune=generic -c init.c -o init.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/gcc  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -std=gnu99 -march=x86-64 -mtune=generic -c ole.c -o ole.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/gcc  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -std=gnu99 -march=x86-64 -mtune=generic -c xls.c -o xls.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/gcc  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -std=gnu99 -march=x86-64 -mtune=generic -c xlstool.c -o xlstool.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++  -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/include" -DNDEBUG -Iwindows -I. -D__USE_MINGW_ANSI_STDIO -I"%LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/Rcpp/include"   -I"%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/include"     -O2 -Wall  -march=x86-64 -mtune=generic -c zip.cpp -o zip.o
%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/bin/g++ -shared -s -o readxl.dll tmp.def RcppExports.o XlsWorkBook.o XlsWorkSheet.o XlsxWorkBook.o XlsxWorkSheet.o endian.o init.o ole.o xls.o xlstool.o zip.o -lRiconv -L%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/lib/x64 -L%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/../Library/mingw-w64/lib -L%LOCALAPPDATA%/CONTIN~1/ANACON~1/envs/py36r/R/bin/x64 -lR
installing to %LOCALAPPDATA%/Continuum/anaconda3/envs/py36r/R/library/readxl/libs/x64
** R
** inst
** preparing package for lazy loading
** help
*** installing help indices
  converting help for package 'readxl'
    finding HTML links ... done
    cell-specification                      html
    finding level-2 HTML links ... done

    excel_sheets                            html
    read_excel                              html
    readxl-package                          html
    readxl_example                          html
** building package indices
** installing vignettes
** testing if installed package can be loaded
* DONE (readxl)
>

hope this helps!

[zaphot]

thanks for the answer! Unfortunately this didn't work. Therefore, i created a new issue: #9184