rstudio

Introduction

The rstudio container runs (as the name already implies) RStudio as a frontend. We have included R as well as Python in the container. RStudio can be used for both programming languages. We also install some important packages

• to access the mongodb (R: mongolite; Python: pymongo)
• for data wrangling (R: tibble, dpyr, jsonlite; Python: numpy, pandas)
• for text-analysis (R: quanteda, spacyr; Python: spacy and models en_core_web_sm, de_core_news_sm)

This list is non-inclusive; some more R packages are available. Please see the dockerfile for details. In general, R packages can be installed manually by using the regular install.packages() command; similar procedure applies for Python packages. Alternatively, add package names either in dockerfile (for R) or in file requirements.txt (for Python). Then rebuild container. Note that the size of container increases with each additional package.

The RStudio instance is available at http://rstudio.servername. Username and password can be set in .env file.

With the start of the container, a preference file (config/rstudio-prefs.json) is loaded automatically in RStudio - including the path of the Python installation. To change preferences, either change file directly or make changes in RStudio GUI; these changes are stored to the preference file automatically (even when the container is stopped).

Also, the volume 01_volume is mounted to the container; scripts within this folder are stored when the container is stopped.

Technical Framework

RStudio is taken from docker rocker repository. Python is added manually in dockerfile.

Main Contributor

Felix Süttmann