R version in notebook kernel is different than R version specified in image tag on M1 Mac

[juskaiser]

What docker image you are using?
jupyter/r-notebook:r-4.1.1

What complete docker command do you run to launch the container (omitting sensitive values)?
docker run -p 8888:8888 jupyter/r-notebook:r-4.1.1

Running on an Apple MacBook Pro Laptop:

• OS: 12.0.1
• Chip: Apple M1 Pro

Docker version: Docker Desktop 4.2.0 (70708)

To Replicate

1. Run docker run -p 8888:8888 jupyter/r-notebook:r-4.1.1 from terminal.
2. From within Jupiter in the running container, create a new terminal window. Run the command R:

R version 3.6.3 (2020-02-29) -- "Holding the Windsock"
Copyright (C) 2020 The R Foundation for Statistical Computing
Platform: aarch64-conda-linux-gnu (64-bit)
...

3. Alternatively in an R-kernal notebook run the command version.

platform       aarch64-conda-linux-gnu     
arch           aarch64                     
os             linux-gnu                   
system         aarch64, linux-gnu          
status                                     
major          3                           
minor          6.3                         
year           2020                        
month          02                          
day            29                          
svn rev        77875                       
language       R                           
version.string R version 3.6.3 (2020-02-29)
nickname       Holding the Windsock  

What do you expect to happen?

I'm expecting that the version of R in the notebook kernel will be the same as the version of R specified in the image tag. In this case I am expecting the version of R in the notebook to be 4.1.1 because the jupyter/r-notebook:r-4.1.1 image tag is being used.

What actually happens?

I am getting an older version of R (3.6.3), as shown in the output above. This is problematic because I'm unable to install any additional R libraries that require R 4.0 or greater.

[mathbunnyru]

@juskaiser the reason for this is explained in #1401

Unfortunately, there is no easy solution, but I hope one day our support for arm images will be much better than it is right now.
I will close this issue as a duplicate.