Open, simple, reusable codebooks for coded datasets.
Open Codebooks is not just a package — it’s a pattern.
The idea is that codebooks should be as open and portable as the data they describe.
- Hash URIs: Every category gets a stable, citable identifier.
- Minimal: Define codebooks in a plain CSV or data frame.
- Templated: Render them into human-readable HTML and machine-readable Turtle (SKOS).
- Static: Publish on GitHub Pages, GitLab Pages, or any static host. No server config.
This repository holds the templates and workflow definition.
The first implementation is in R, via the rOpenCodebooks package.
But the design is language-agnostic: we hope to see implementations in Python, Julia, and other environments where people work with coded datasets.
Surveys, statistics, and research data often use categorical variables.
To make these datasets comparable and FAIR, the categories must be:
- Clearly defined
- Linked to stable identifiers
- Published in reusable formats
Open Codebooks makes this easy — for one question, one dataset, or across a whole survey program.
- ✅ Reference implementation in R
- 🔜 Ports to Python and Julia
- 🔜 Community registry of common codebooks (e.g. agreement scales, trust scales, SDMX extensions)
MIT — free to use, adapt, and port.