SliTaz Forum
You are not logged in.
Pages: 1
Hello.
I've tryed to install R under Slitaz 5.0 and noticed TWO bugs
1- the first is about install (tazpkg with browser interface) : when you type "R" in the panel, the processor speed up to 100%, but, finally I can install R. I sippose that the package manager cheks all packages with the "R" letter ...
2- very annoying bug : "R" can't start, because, under the terminal (R is a terminal based sodtware), I can see that there is a shared library that lacks in Slitaz :
libquadmath.so.0
So, for me, the packager should very quickly checks dependencies, because "R" is a very, very useful package for researchers.
Regards.
Offline
Install one of these packages:
http://pkgs.slitaz.org/?file=libquadmath.so.0
Offline
Thanks Ceel.
I've installed the last Gcc package. In "R" starts now.
But for me, the "R" package must be corrected in order to install the devlopment tools like Gcc. As you know, "R" extensions are often supplyed as source.
And, for now, the "R" package in Slitaz as only Fortran as dependencie ... not C/C++
Regards.
Offline
and for compiling within "R" itself, one must install the "slitaz-toolchain" and also "bash"
Offline
In order to avoid opening a new post, I'm happy to informe you that I've successly installed R 3.4 under Slitaz 5.0
You have to get the source package from R-Cran, and of course, compile it.
And of course R 3.4 requires some extra packages from Slitaz repositories :
1- the Slitaz tool chain in order to get Gcc, make, and so on ... in other words, the basic compilation tools
2- the gfortran/fortran extensions (see Slitaz package manager)
3- zma/libzma (the library itself, and the corresponding .dev package)
4- curl/libcurl (" """ ")
5- readline ( " ")
6- optinnal : because i've done some tricky job before, TCL/TK (programs + dev files) but I'm not sure if it's required.
7- expand the source file code in your /home directory (for example)
8- open a Terminal where you have decompressed the archive
9- type ./configure en hit <return> key. If all dependencies, the configuration stops without error(s) in other cases you have to get and install missing software and restart ./configure (step 9)
10- when all is OK, type make and hit <return> key, and then you can try a walk until the end of compilation (45 minutes in my old but pretty Akoya netbook).
11- now the dirty/tricky phase : type sudo make install and hit <return>. The install is OK, R works, but the installation is incorrect for Slitaz. You must buikld a symbolic link between the /usr/bin directory, where R is supposed to be, and the true directory where R is installed, e.g. /usr/local/bin/R
Or, you can write a corrected R laucher in /usr/bin (with root permissions) ; here is the code. Save it in /usr/bin under R (please note uppercase) and make it executable.
Here is the file ;
Here is a screenshot of my R 3.4 running under Slitaz 5.0
Offline
EDIT : I've found that there is an entry in Slitaz main menu, under /science ; very useful and better than under Ubuntu !
Offline
And, I've done more tests, I've installed a heavy package, the "R commander " extension, a graphical frontend for "R". It's one of my reasons to upgrade from the Slitaz package (3.1) without access to tcl/tk and the current 3.4 version with this support.
After again 45 minutes of downloading and compilation, I've get a fonctionnal "Rcmdr" extension :
[attachment=44784,2508]
Offline
I've also installed the "rtemis" plugin for Rcmdr ; "Rtemis" is a cool statistical tool to build deep text analysis for social science research. It's a free extension for Rcmdr. Here you can see a screenshot with a factorial analysis :
Rtemis is a part of official R-Cran repositories. You can install it from within R with this
install.packages("RcmdrPlugin.temis")
[attachment=44785,2509]
Offline
It's possible to get, convert and install RStudio version 0.98 and use it under Slitaz with R.
Here is the URL (functionnal in June 2017) to get ol versions of R-Studio :
https://support.rstudio.com/hc/en-us/articles/206569407-Older-Versions-of-RStudio
But, please note : in order to use a self-compiled version of "R" with RStudio, you must start the compilation with this instruction :
./configure --enable-R-shlib
And, there is some trouble in the lauching method of RStudio ! You must tell to RStudio Where is you "R" laucher.
Here is an example :
RSTUDIO_WHICH_R=/home/luc/r-cran/R-3.4.0/bin/R /usr/lib/rstudio/bin/rstudio
Another issue : in order to get all graphics capabilities of "R", you MUST add a .Rprofile hidden file, for example in your /home directory
I Can't add any .txt file, I don't kow why, so here is the content of this .RProfile file :
# default X11() setting
setHook(packageEvent("grDevices", "onLoad"),
function(...) grDevices::X11.options(width = 8, height = 8, xpos = 0,pointsize = 10, type = "cairo"))
## Default repo
local({r <- getOption("repos")
r["CRAN"] <- "http://cran.r-project.org"
options(repos=r)
})
Offline
Pages: 1
[ Generated in 1.635 seconds, 7 queries executed - Memory usage: 1.55 MiB (Peak: 1.77 MiB) ]