Update for RStudio Server and R/4.2.1

config.yaml

@@ -59,14 +59,15 @@ contact: 'grosscra@msu.edu'
 
 # Order of episodes in your lesson
 episodes:
-- rstudio-ondemand.Rmd
+- rstudio-server-ondemand.Rmd
 - managing-r-env-packages.Rmd
 - parallelizing-r-code.Rmd
 - r-command-line.Rmd
 - r-slurm-jobs.Rmd
 
 # Information for Learners
 learners:
+- rstudio-ondemand.Rmd
 
 # Information for Instructors
 instructors:

episodes/managing-r-env-packages.Rmd

@@ -31,15 +31,15 @@ We can do this by typing `.libPaths()` into the R console:
 .libPaths()
 ```
 ```output
-[1] "/mnt/ufs18/home-237/k0068027/R/x86_64-pc-linux-gnu-library/4.0"
-[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.0.3-foss-2020a/lib64/R/library" 
+[1] "/mnt/ufs18/home-237/k0068027/R/x86_64-pc-linux-gnu-library/4.2"
+[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.2.1-foss-2022a/lib64/R/library" 
 ```
 
 We see two directories.
 The first is created for you in your home directory, and the second (or one like it, starting with `/cvmfs/pub_software.icer.msu.edu/software` or `/opt/software`) points to all of the packages that are pre-installed on the HPCC.
 When you use `install.packages()` in the future, by default, it will install to the first entry in your `.libPaths()`.
 
-One important point to note is that the library in your home directory is labeled with `4.0` for version 4.0(.3) of R.
+One important point to note is that the library in your home directory is labeled with `4.2` for version 4.2(.1) of R, the default used by RStudio Server.
 If you ever use different versions of R, it is important that the packages you use are consistent with those versions.
 So, for example, if you choose to use R/3.6.2, you should make sure that the library in your home directory returned by `.libPaths()` ends in 3.6.
 Mixing versions will likely cause your packages to stop working!
@@ -65,7 +65,7 @@ Luckily, R knows this, and if you try to install a package, you will be offered
 
 ``` output
 Warning in install.packages :
-  'lib = "/cvmfs/pub_software.icer.msu.edu/software/R/4.0.3-foss-2020a/lib64/R/library"' is not writable
+  'lib = "/cvmfs/pub_software.icer.msu.edu/software/R/4.2.1-foss-2022a/lib64/R/library"' is not writable
 Would you like to use a personal library instead? (yes/No/cancel)
 ```
 
@@ -122,7 +122,7 @@ Here are some general tips:
 
 1. Read the documentation for the package you're using and take note of any dependencies you need and their versions. This information is also included under SystemRequirements on a package's [CRAN](https://cran.r-project.org/) page.
 2. Make sure that software is available before you try to install/use the R package. This could involve:
-    - Loading it through the HPCC module system. **Note**: This is not possible (yet) in RStudio on the HPCC. You will have to [use R through the command line](r-command-line.Rmd#loading-external-dependencies).
+    - Loading it through the HPCC module system. To do this in OnDemand, click the Advanced Options" checkbox when you start a new RStudio Server session. The first option will allow you to enter HPCC modules you'd like to load before RStudio starts. Otherwise, you can [load these packages and use R through the command line](r-command-line.Rmd#loading-external-dependencies).
     - Installing it yourself in a way that R can find it.
 3. If a package's setup instructions suggest something like `sudo apt-get ...` or `sudo dnf install ...` under the Linux instructions, this is a sign that it needs external dependencies.
 These methods won't work for installation on the HPCC; instead, look for and load HPCC modules with similar names.
@@ -186,9 +186,9 @@ To make sure this runs every time we start R, we'll put it in the `.Rprofile` fi
 
 Use RStudio to open a new Text File and type
 
-```text
+```r
 local({
-options(repos = c(CRAN="https://repo.miserver.it.umich.edu/cran/"))
+  options(repos = c(CRAN="https://repo.miserver.it.umich.edu/cran/"))
 })
 ```
 
@@ -197,7 +197,15 @@ It's good practice to put any code you write in your `.Rprofile` in a call to `l
 
 Save this in your `r_workshop` directory as `.Rprofile` (don't forget the leading `.`).
 Any time R starts, it will look for a `.Rprofile` file in the current directory, and execute all of the code before doing anything else.
-To make this take effect in RStudio, you can restart R by going to the Session menu, and select Restart R.
+To make this take effect in RStudio, you can restart R by going to the Session menu, and select Restart R. To check our work, run
+
+```r
+options()$repos
+```
+```output
+                                      CRAN 
+"https://repo.miserver.it.umich.edu/cran/"
+```
 
 Now suppose that this project we're working on uses some very special packages that we don't want in the library in our home directory.
 The right way to do this is with a package manager like [`packrat`](https://rstudio.github.io/packrat/) or the newer [`renv`](https://rstudio.github.io/renv/articles/renv.html).
@@ -229,7 +237,7 @@ Now, restart R using the Session menu, and check your `.libPaths()` in the R con
 ```
 ```output
 [1] "/mnt/ufs18/home-237/k0068027/r_workshop/library" 
-[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.0.3-foss-2020a/lib64/R/library"
+[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.2.1-foss-2022a/lib64/R/library"
 ```
 
 Great! We can even check that we've isolated ourselves from the default home directory library by trying to load `cowsay`:
@@ -283,7 +291,7 @@ Double checking our library paths
 ```
 ```output
 [1] "/mnt/ufs18/home-237/k0068027/r_workshop/library" 
-[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.0.3-foss-2020a/lib64/R/library"
+[2] "/cvmfs/pub_software.icer.msu.edu/software/R/4.2.1-foss-2022a/lib64/R/library"
 ```
 
 we see that our `r_workshop/library` directory is first.

episodes/r-command-line.Rmd

@@ -77,42 +77,40 @@ module spider R
       $ module -r spider '.*R.*'
 ```
 
-We've abbreviated the output, but we can see that there are lots of different versions of R available! We'll try loading 4.0.3 since that version on the HPCC has a large number of packages pre-installed.
+We've abbreviated the output, but we can see that there are lots of different versions of R available! We'll try loading 4.2.1 since that version matches the one used in the RStudio Server OnDemand app.
 
 If you're familiar with the module system, you might try to load the module right away with `module load`:
 
 ```bash
-module load R/4.0.3
+module load R/4.2.1
 ```
 
 ```output
 Lmod has detected the following error:  These module(s) or extension(s) exist
-but cannot be loaded as requested: "R/4.0.3"
-   Try: "module spider R/4.0.3" to see how to load the module(s).
+but cannot be loaded as requested: "R/4.2.1"
+   Try: "module spider R/4.2.1" to see how to load the module(s).
 ```
 
 But we get an error! Let's try the suggested fix to see what's going on:
 
 
 ```bash
-module spider R/4.0.3
+module spider R/4.2.1
 ```
 
 ```output
 ----------------------------------------------------------------------------
-  R: R/4.0.3
+  R: R/4.2.1
 ----------------------------------------------------------------------------
     Description:
       R is a free software environment for statistical computing and
       graphics.
 
 
     You will need to load all module(s) on any one of the lines below before the
-    "R/4.0.3" module is available to load.
+    "R/4.2.1" module is available to load.
 
-      GCC/10.2.0  OpenMPI/4.0.5
-      GCC/9.3.0  OpenMPI/4.0.3
-      iccifort/2020.1.217  impi/2019.7.217
+      GCC/11.3.0  OpenMPI/4.1.4
  
     Help:
       Description
@@ -129,7 +127,7 @@ Before we do that, it's good practice to purge any other modules that might be l
 
 ```bash
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 ```
 
 No error! Let's check that we can access R:
@@ -139,8 +137,8 @@ R
 ```
 
 ```output
-R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
-Copyright (C) 2020 The R Foundation for Statistical Computing
+R version 4.2.1 (2022-06-23) -- "Funny-Looking Kid"
+Copyright (C) 2022 The R Foundation for Statistical Computing
 Platform: x86_64-pc-linux-gnu (64-bit)
 
 R is free software and comes with ABSOLUTELY NO WARRANTY.
@@ -158,6 +156,7 @@ Type 'demo()' for some demos, 'help()' for on-line help, or
 Type 'q()' to quit R.
 
 > 
+
 ```
 
 Great! We now have an R console where we can run short lines of code, just like from RStudio. As the output from `R` shows, type `q()` to quit and return back to the command line. 
@@ -174,36 +173,37 @@ We will use this option below to make sure we run our code in the cleanest envir
 When you load the R module using the `module load` commands above (before you actually run `R`), this is also the time to load those external dependencies.
 
 Note that these dependencies and R will all need to be be compatible (e.g., use the same version of GCC and MPI).
-For example, a spatial ecology workflow might require the use of GDAL and UDUNITS as dependencies for R packages.
+For example, a Bayesian modeling workflow might require the use of JAGS as dependencies for R packages.
 After loading R and its dependencies with
 
 ```bash
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 ```
 
 you can try loading a compatible GDAL and UDUNITS without specifying a version:
 
 ```bash
-module load GDAL UDUNITS
+module load JAGS
 ```
 
-Then check which versions got loaded:
+Then check which version gets loaded:
 
 ```bash
 module list
 ```
 ```output
 Currently Loaded Modules:
-  1) GCCcore/10.2.0      27) zstd/1.4.5           53) FLAC/1.3.3
-  2) zlib/1.2.11         28) libdrm/2.4.102       54) libvorbis/1.3.7
+  1) GCCcore/11.3.0      34) zstd/1.5.2           67) ICU/71.1
+  2) zlib/1.2.12         35) libdrm/2.4.110       68) Szip/2.1.1
+  3) binutils/2.38       36) libglvnd/1.4.0       69) HDF5/1.12.2
  ...
- 24) X11/20201008        50) GMP/6.2.0            76) GDAL/3.3.2
- 25) gzip/1.10           51) NLopt/2.6.2          77) UDUNITS/2.2.26
- 26) lz4/1.9.2           52) libogg/1.3.4
+ 31) X11/20220504        64) libopus/1.3.1        97) PROJ/9.0.0
+ 32) gzip/1.12           65) LAME/3.100           98) JAGS/4.3.1
+ 33) lz4/1.9.3           66) libsndfile/1.1.0
 ```
 
-If these versions will work, then great! 
+If this version will work, then great! 
 If not, then you might try using a different version of R.
 Usually, newer versions of dependencies for popular R packages get installed with newer versions of R.
 

episodes/r-slurm-jobs.Rmd

@@ -65,7 +65,7 @@ This code is exactly what you would enter on the command line to run your R scri
 
 # Load the R module
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 
 # Get to our project directory
 cd ~/r_workshop
@@ -125,12 +125,12 @@ JobId=17815750 JobName=single_core.sh
    MinCPUsNode=1 MinMemoryCPU=500M MinTmpDiskNode=0
    Features=[intel14|intel16|intel18|(amr|acm)|nvf|nal|nif] DelayBoot=00:00:00
    OverSubscribe=OK Contiguous=0 Licenses=(null) Network=(null)
-   Command=/mnt/ufs18/home-045/grosscra/r_workshop/slurm/single_core.sh
-   WorkDir=/mnt/ufs18/home-045/grosscra/r_workshop
-   Comment=stdout=/mnt/ufs18/home-045/grosscra/r_workshop/slurm-17815750.out 
-   StdErr=/mnt/ufs18/home-045/grosscra/r_workshop/slurm-17815750.out
+   Command=/mnt/ufs18/home-237/k0068027/r_workshop/slurm/single_core.sh
+   WorkDir=/mnt/ufs18/home-237/k0068027/r_workshop
+   Comment=stdout=/mnt/ufs18/home-237/k0068027/r_workshop/slurm-17815750.out 
+   StdErr=/mnt/ufs18/home-237/k0068027/r_workshop/slurm-17815750.out
    StdIn=/dev/null
-   StdOut=/mnt/ufs18/home-045/grosscra/r_workshop/slurm-17815750.out
+   StdOut=/mnt/ufs18/home-237/k0068027/r_workshop/slurm-17815750.out
    Power=
 ```
 
@@ -159,7 +159,7 @@ Submit the job and compare the time it took to run with the single core job.
 
 # Load the R module
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 
 # Get to our project directory
 cd ~/r_workshop
@@ -220,7 +220,7 @@ For the steps below, you will need the [list of SLURM job specifications](https:
 
 # Load the R module
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 
 # Get to our project directory
 cd ~/r_workshop
@@ -275,7 +275,7 @@ Here is an example script:
 
 # Load the R module
 module purge
-module load GCC/10.2.0 OpenMPI/4.0.5 R/4.0.3
+module load GCC/11.3.0 OpenMPI/4.1.4 R/4.2.1
 
 # Get to our project directory
 cd ~/r_workshop