Code Reproducibility Good Practices

coding_good_practices.md

R Code Reproducibility Good Practices

Minimal Rules

1. Organize Your R Analysis Using Projects in Rstudio
   • Ensure at least three folders exist: code, data, and figures.
   • Tip: use analysistemplates package

     remotes::install_github("jonas-hag/analysistemplates")

2. Use {here} and Avoid setwd
   • Never use setwd, ever.
3. Divide Analysis into Specific Scripts
   • Use at least 01_data_acquisition_cleaning and 02_exploratory_data_analysis. Create additional scripts for modeling as needed.
4. Store Data Frames as RDS in data Folder
   • The output of file 01 should be a data frame, preferably stored as an RDS for retaining order and metadata. If no relevant metadata/ordered levels are in the dataframe, a csv file will suffice.
5. Maintain a Clutter-Free Environment Tab
   • Promptly remove unnecessary objects.
6. Include Package References at the Beginning
   • Use package::function syntax for specific function calls.
7. Comment Everything and Everywhere
   • Never assume you will remember your reasoning or logic in the future.
8. Adhere to the Google R Style Guide
   • Regularly apply Control+Shift+A for automatic formatting.

Reference materials

• Computational Reproducibility
• The Turing Way handbook to reproducible, ethical and collaborative data science
• A Guide to Reproducible Code in Ecology and Evolution
• Putting the R into Reproducible Research