Takeaway
This guide records how I end up:
- use R (and other Linux tools) on Linux via WSL
- use Rstudio via any browser on Windows without leaving Unix platform
- use computation library (blas/lapack) for R
What you need
I use Windows Terminal to run my command line.
Before going through this guide, make sure you have your WSL installed, see document.
You can simply run the script as administrator if you never install WSL:
wsl --install
I installed the default Ubuntu using following script. It installed Ubuntu 20.04 LTS for me. You can install other distro both I like the default one.
wsl --install -d Ubuntu
and make sure I have the right version:
$ wsl -l -v
NAME STATE VERSION
* Ubuntu Running 2
Set up Ubuntu
I still want 22.04 version! So I upgrade Ubuntu to 22.04 LTS. -d flag will upgrade to development version. You don’t have to but the script of installing Rstudio will be a bit different.
sudo do-realse-upgrade
sudo apt update
sudo apt upgrade
Install R core and add CRAN repository
I followed this guide to install R and incorporate CRAN into apt repository:
# update indices
sudo apt update -qq
# install two helper packages we need
sudo apt install --no-install-recommends software-properties-common dirmngr
# add the signing key (by Michael Rutter) for these repos
# To verify key, run gpg --show-keys /etc/apt/trusted.gpg.d/cran_ubuntu_key.asc
# Fingerprint: E298A3A825C0D65DFD57CBB651716619E084DAB9
wget -qO- https://cloud.r-project.org/bin/linux/ubuntu/marutter_pubkey.asc | sudo tee -a /etc/apt/trusted.gpg.d/cran_ubuntu_key.asc
# add the R 4.0 repo from CRAN -- adjust 'focal' to 'groovy' or 'bionic' as needed
sudo add-apt-repository "deb https://cloud.r-project.org/bin/linux/ubuntu $(lsb_release -cs)-cran40/"
sudo apt install --no-install-recommends r-base
sudo add-apt-repository ppa:c2d4u.team/c2d4u4.0-
Let’s see if R is there:
$ R
> sessionInfo()
R version 4.2.1 (2022-06-23)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 22.04.1 LTS
Install RStudio Server
I installed RStudio Server because I can use Rstudio via browser, which makes no difference. If you want to use RStudio Desktop, you will need GUI software to make it works unless you have OS Build higher than 10.0.22000, which support native GUI. But why bother?
I followed this guide to install RStudio for Ubuntu. It is a bit different for Ubuntu 20 v.s. 22.
sudo apt install -y r-base r-base-core r-recommended r-base-dev gdebi-core build-essential libcurl4-gnutls-dev libxml2-dev libssl-dev
wget https://download2.rstudio.org/server/jammy/amd64/rstudio-server-2022.07.1-554-amd64.deb
sudo gdebi rstudio-server-2022.07.1-554-amd64.deb
After this, Rstudio server automatically starts. In my terminal, it drops out something. Don’t panic:
TTY detected. Printing informational message about logging configuration. Logging configuration loaded from '/etc/rstudio/logging.conf'. Logging to '/var/log/rstudio/rstudio-server/rserver.log'.
Now go to localhost:8787 and enter your username and password for Ubuntu!
Install BLAS and LAPACK to accelerate your matrix computation
BLAS and LAPACK are useful because it improves R performance, see the article.
Install BLAS and LAPACK on Windows
Loading BLAS and LAPACK libraries for R on Windows is difficult (at least to me). The most “popular” way to directly overwrite the R math library (Rblas.dll and Rlapack.dll). Brutal, but it works.
Until I update my R and everything gone. I have to redo it. Also, it is annoying when I tried to use an R package (semPlot) and it dropped off error message. There is something wrong, I know. But I couldn’t figure it out at that time. I went back to R’s default library.
Install BLAS and LAPACK on Linux
It seems more straghtforward on Linux, because it is easier to complie your own R core by passing some files and flags like --with-blas. But I am not knowledgable enough to handle this. Also, some tutorials suggest use dynamic link, but it does not work for me.
Here is the way I realize how to make everything works well.
R-core without default Rblas & Rlapack
First, I realize that in my R on Linux, there is no file libRblas.so and libRblas.so. There is only one file which once I changed it, erros occur and my R process shut down.
$ ls /usr/lib/R/lib/
libR.so
It leads me to a stackoverflow question, where [Dirk Eddelbuettel(https://dirk.eddelbuettel.com), the guru of R, explained how they design the R core. In a simple way, R installed in my way is not built with libraries it carries. Therefore, my R is open to any system-wide library.
“allows you to switch BLAS installation”
If we look into R, it shows the default matrix libraries are:
> sessionInfo()
Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.10.0
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.10.0
Now let swtich installation for R. All we need to do is install BLAS & LAPACK libraries, and they will automatically become system-wide libraries. Maybe some people don’t like it, but I don’t care.
Intel oneAPI Math Kernel Library (oneMKL)
You can choose whatever library you want, I use [Intel oneAPI Math Kernel Library(https://www.intel.com/content/www/us/en/developer/tools/oneapi/onemkl.html#gs.9dcm00) for my office computer, and OpenBlas for my PC.
# Install Intel MKL
sudo apt install intel-mkl
# sudo apt install intel-mkl-full
# You may use this if intel-mkl does not make libraries system-wide
# update libraries priority
sudo update-alternatives --config libblas.so.3-x86_64-linux-gnu
sudo update-alternatives --config liblapack.so.3-x86_64-linux-gnu
# each line, it will prompt
Selection Path Priority Status
------------------------------------------------------------
0 /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 100 auto mode
1 /usr/lib/x86_64-linux-gnu/blas/libblas.so.3 10 manual mode
2 /usr/lib/x86_64-linux-gnu/libmkl_rt.so 1 manual mode
* 3 /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 100 manual mode
Press <enter> to keep the current choice[*, or type selection number:
# type the number where libmkl_rt.so at
# !remember there are two lines that you should run
Going back to R, you will see
$ R
> sessionInfo()
Matrix products: default
BLAS/LAPACK: /usr/lib/x86_64-linux-gnu/libmkl_rt.so
Also, it seems that we need to run additional script according to Intel Document and many tutorials. I am not sure if it is necessary to do so. Maybe you can run it anyway:
export MKL_INTERFACE_LAYER=GNU,LP64
export MKL_THREADING_LAYER=GNU
OpenBlas
You can install OpenBlas (or other BLAS/LAPACK libraries like ATLAS) in a similar way:
# Install Intel MKL
sudo apt install libopenblias-dev
# sudo apt install libopenblias-base
# You may use this, but I don't know the differences
# For details, see https://askubuntu.com/questions/1318554/why-are-there-so-many-openblas-packages-and-which-one-would-yield-fastest-result
# update libraries priority
sudo update-alternatives --config libblas.so.3-x86_64-linux-gnu
sudo update-alternatives --config liblapack.so.3-x86_64-linux-gnu
# each line, it will prompt
Selection Path Priority Status
------------------------------------------------------------
* 0 /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 100 auto mode
1 /usr/lib/x86_64-linux-gnu/blas/libblas.so.3 10 manual mode
2 /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 100 manual mode
Press <enter> to keep the current choice[*, or type selection number:
# type the number where libblas.so.3 and liblapack.so.3 at
# !remember there are two lines that you should run
Going back to R, you will see
$ R
> sessionInfo()
Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.20.so
Other Approach
It seems there is a wrapper library flexiblas allowing switch among different BLAS/LAPACK libraries. There is also a R package, “Flexiblas” using API by the wrapper library.
I may take a look later but current solution works pretty well.
System Info
Windows:
Processor: AMD Ryzen 5 5600H with Radeon Graphics
PSVersion 7.2.5
PSEdition Core
GitCommitId 7.2.5
OS Microsoft Windows 10.0.22000
Platform Win32NT
PSCompatibleVersions {1.0, 2.0, 3.0, 4.0…}
PSRemotingProtocolVersion 2.3
SerializationVersion 1.1.0.1
WSManStackVersion 3.0
WSL:
WSL version: 0.65.3.0
Kernel version: 5.15.57.1
WSLg version: 1.0.41
MSRDC version: 1.2.3213
Direct3D version: 1.601.0
DXCore version: 10.0.25131.1002-220531-1700.rs-onecore-base2-hyp
Windows version: 10.0.22000.856
Ubuntu:
Distributor ID: Ubuntu
Description: Ubuntu 22.04.1 LTS
Release: 22.04
Codename: jammy