Topic 8 | Course Wrap-up
Date: May, 2026
Synopsis: Final session to complete pending work, review key concepts, and ensure every project is in a reproducible state.
Class overview: By the end of this class you will have: resolved any remaining issues in your project, reviewed the core principles of the course, and pushed a final, complete version of your repository to GitHub.
Two groups (Master’s students and PhD students) run independently, same structure.
| Segment | Duration |
|---|---|
| Final completeness check | 20 min |
| Finish pending work | 40 min |
| Review of key concepts | 40 min |
| Final commit and closing | 20 min |
| Total | ~120 min |
Final Completeness Check
Pull your project and verify that everything is in place. Use this checklist:
The instructor will review each student’s checklist and flag any critical gaps.
If your Quarto report does not render from a clean R session, that is the first thing to fix. A report that only works on your machine is not reproducible.
Finish Pending Work
Use this time to close any remaining gaps identified in the checklist above. The instructor circulates to help. Prioritise items that affect reproducibility — a broken rendering pipeline or an undocumented variable matters more than a plot caption.
Review of Key Concepts
A brief walkthrough of the core ideas that tie the course together.
The reproducibility chain
Everything in this course serves one principle: someone who has never seen your project should be able to clone the repository, read the README, run the code, and obtain the same results you did. Each element you built supports this chain:
- Raw data is preserved untouched. All transformations happen in code.
- The tidying script turns raw data into an analysis-ready format, and the transformation is documented and repeatable.
- The Quarto report combines the analysis with its narrative, so the reasoning is as reproducible as the computation.
- The data descriptor documents the dataset independently of the analysis, making it understandable and reusable.
- Git tracks every change. GitHub makes the project accessible.
Tidy data
One variable per column. One observation per row. One value per cell. This structure is not arbitrary — it follows from the experimental design and determines what analyses are possible. Identifying the observational unit is the first and most important step.
Version control
Git records what changed, when, and why. Meaningful commit messages are documentation. The habit of committing small, coherent changes makes it possible to trace any result back to the code that produced it.
Literate programming
Code without explanation is hard to trust. Prose without code is hard to verify. A Quarto report provides both, rendered from a clean session, ensuring that what the reader sees is what the code actually produces.
Data documentation
A dataset without a description is a dataset that cannot be evaluated or reused. The data descriptor — inspired by Nature Scientific Data — provides the metadata, variable definitions, and quality evidence that make a dataset a proper scientific product.
Final Commit and Closing
Commit and push
- Stage all remaining changes.
- Write a final commit message:
"Final project state for course submission". - Push to GitHub.
Verify on GitHub
Open your repository on GitHub and confirm that the files are there and up to date. This is what the instructor (and any future reader) will see.
All course materials remain available at instructr.github.io/ualg.rep.data.analysis.2026 for future reference.