Dockerized Business Analytics

This repo contains information to setup a docker image with R, Rstudio, Shiny, Radiant, Python, Postgres, JupyterLab, and Code-Server (aka VSCode)

Install Docker

To use the docker images you first need to install Docker

• For Mac: https://docs.docker.com/docker-for-mac/
• For Windows: https://docs.docker.com/docker-for-windows/
• For Linux: https://docs.docker.com/engine/installation/

After installing Docker, check that it is running by typing docker --version in a terminal. This should return something like the below:

docker --version
Docker version 18.09.2, build 6247962

On windows please install Git Bash:

http://www.techoism.com/how-to-install-git-bash-on-windows/

For detailed install instructions on Windows see install/rsm-msba-windows.md

For detailed install instructions on macOS see install/rsm-msba-macos.md

r-focal

You probably don't want to run this image by itself. It is used in the radiant, rsm-msba-spark, and rsm-jupyterhub, application (see below). To build a new container based on r-focal add the following at the top of your Dockerfile

FROM vnijs/r-focal:latest

To build r-focal yourself use:

docker build -t $USER/r-focal ./r-focal

To push to docker hub use:

sudo docker login 
docker push $USER/r-focal

radiant

The second image builds on r-focal and adds radiant and required R-packages. To build a new container based on radiant add the following at the top of your Dockerfile

FROM vnijs/radiant:latest

To allow execution of R-code in Report > Rmd and Report > R in Radiant add the following to .Rprofile in your home directory

options(radiant.ace_vim.keys = FALSE)
options(radiant.maxRequestSize = -1)
# options(radiant.maxRequestSize = 10 * 1024^2)
options(radiant.report = TRUE)
# options(radiant.shinyFiles = TRUE)
# options(radiant.ace_theme = "cobalt")
options(radiant.ace_theme = "tomorrow")
# options(radiant.ace_showInvisibles = TRUE)

rsm-msba and rsm-msba-spark

The third and forth images build on the radiant image and adds python, jupyter lab, postgresql, spark, and Code-Server (aka VSCode). To build a new container based on rsm-msba-spark add the following at the top of your Dockerfile

FROM vnijs/rsm-msba-spark:latest

rsm-jupyterlab

This image builds on rsm-msba-spark and is set up to be accessible from a server running jupyter hub.

rsm-vscode

This image contains all R and Python libraries found in rsm-msba-spark and rsm-jupterlab but does not include Jupyter Lab, Shiny server, Rstudio server, or VSCode (codeserver). It is intended to be used with a local install of VSCode. Once you start the container using launch-rsm-vscode you can use "Remote-Containers: Attach to Running Container" from VSCode to connect. Recommended extensions to use with the container are shown in the screenshot below:

Trouble shooting

To stop (all) running containers use:

docker kill $(docker ps -q)

If the build fails for some reason you can access the container through the bash shell using to investigate what went wrong:

docker run -t -i $USER/rsm-msba-spark /bin/bash

To remove an existing image use:

docker rmi --force $USER/rsm-msba-spark

To remove stop all running containers, remove unused images, and errand docker processes use the dclean.sh script

./scripts/dclean.sh

General docker related commands

Check the disk space used by docker images

docker ps -s

docker system df

Trademarks

Shiny and Shiny Server are registered trademarks of RStudio, Inc. The use of the trademarked terms Shiny and Shiny Server and the distribution of the Shiny Server through the images hosted on hub.docker.com has been granted by explicit permission of RStudio. Please review RStudio's trademark use policy and address inquiries about further distribution or other questions to permissions@rstudio.com.

Jupyter is distributed under the BSD 3-Clause license (Copyright (c) 2017, Project Jupyter Contributors)

Acknowledgements

Thanks to Ajar Vashisth for helping me get started with Docker and Docker Compose