R in Jupyter

Jupyter allows for language kernels other than Python.

R version

There two versions of R on Picotte. Pick one, and set up your .bashrc file to use that one by default. See the article on R.

For this article, we will use the default R 4.0.2 (without loading a modulefile).

If you want to use R 4.1.3 via the modulefile R/4.1.3, make sure to add this line to your .bashrc file so that it becomes your default R. You will also need to do

    module load R/4.1.3

before you follow any of the following steps.

Prerequisites

VS Code Jupyter setup

This requires that you have VS Code set up to use remote Jupyter kernels. See: Jupyter via VS Code

We will use the following Python version (Jupyter is a Python application; the Jupyter kernel can be other languages):

[juser@picotte001 ~]$ module load python/gcc/3.10

R Packages

For an R kernel, some R packages need to be installed.

roxygen2
devtools
IRkernel[1][2]

At the R command line (i.e. in a terminal in VS Code just type "R")

[juser@picotte001 ~]$ module load R
[juser@picotte001 ~]$ R

R version 4.2.2 (2022-10-31) -- "Innocent and Trusting"
Copyright (C) 2022 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
...
> install.packages("roxygen2")
> install.packages("devtools")
Warning in install.packages("devtools") :
  'lib = "/opt/microsoft/ropen/4.0.2/lib64/R/library"' is not writable
Would you like to use a personal library instead? (yes/No/cancel) yes
Would you like to create a personal library
‘~/R/x86_64-pc-linux-gnu-library/4.0’
to install packages into? (yes/No/cancel) yes
... [some minutes of installation and lots of output] ...
> install.packages("IRkernel")
> IRkernel::installspec(name = 'ir42', displayname = 'R 4.2')   # modify this if you're using a version of R different from 4.2

Then, exit R by typing "Ctrl-D".

N.B.

install.packages() in R 4.1.3 may prompt you to select a mirror site. Choose one in the US close to Philadelphia.
some of these packages may have already been installed globally, so you may see output that indicates that fact.
if any package installation fails, wait a few minutes and try again

Jupyter Lab extension

If using python/gcc/3.10, this should already be installed.

Run Jupyter Lab

Next, run Jupyter Lab. In the terminal:

[juser@picotte001 ~]$ jupyter-lab
[I 2022-09-29 16:09:18.115 ServerApp] jupyterlab | extension was successfully linked.
[I 2022-09-29 16:09:18.124 ServerApp] nbclassic | extension was successfully linked.
[I 2022-09-29 16:09:18.776 ServerApp] notebook_shim | extension was successfully linked.
[I 2022-09-29 16:09:18.841 ServerApp] notebook_shim | extension was successfully loaded.
[I 2022-09-29 16:09:18.841 LabApp] JupyterLab extension loaded from /ifs/opt/python/gcc/3.10/lib/python3.10/site-packages/jupyterlab
[I 2022-09-29 16:09:18.841 LabApp] JupyterLab application directory is /ifs/opt/python/gcc/3.10/share/jupyter/lab
[I 2022-09-29 16:09:18.845 ServerApp] jupyterlab | extension was successfully loaded.
[I 2022-09-29 16:09:18.862 ServerApp] nbclassic | extension was successfully loaded.
[I 2022-09-29 16:09:18.862 ServerApp] Serving notebooks from local directory: /home/rbatty
[I 2022-09-29 16:09:18.862 ServerApp] Jupyter Server 1.18.1 is running at:
[I 2022-09-29 16:09:18.862 ServerApp] http://localhost:8888/lab?token=random_characters
...
[I 2022-09-29 16:09:18.862 ServerApp]  or http://127.0.0.1:8888/lab?token=random_characters
[I 2022-09-29 16:09:18.862 ServerApp] Use Control-C to stop this server and shut down all kernels (twice to skip confirmation).
[C 2022-09-29 16:09:18.882 ServerApp]

    To access the server, open this file in a browser:
        file:///home/rbatty/.local/share/jupyter/runtime/jpserver-14372-open.html
    Or copy and paste one of these URLs:
        http://localhost:8888/lab?token=random_characters
     or http://127.0.0.1:8888/lab?token=random_characters

Copy the the URL shown (containing the token of random characters. Open a new browser window on your PC and paste the URL into the address bar.

From Jupyter Lab's "File" menu, select "New Launcher". You should see a new tab, where you can click on the "R" button under "Notebook":

756px

When you click it, you should have a new notebook using the R kernel.

756px

NOTE

these files reside on Picotte, not on your PC.
the R application runs on `picotte001`, the login node, which is concurrently used by other users. Please do not run large computations in a notebook like this.