Draft Quarto Analysis

Day 4 Morning | EDS 214

Thursday August 28th, 2025

Background

You started this week by looking at an analysis you needed to replicate. Since then, you wrote some spaghetti code and started refactoring your analysis. You took steps towards a collaborative, reproducible analysis by:

  • Organizing your folders and files

  • Tracking tasks using GitHub Issues

  • Isolating changes in branches

But most of your analysis is still in a spaghetti state…

Goals

In this session, you’re going to convert your spaghetti code into a reproducible Quarto document. Then you’ll use GitHub Pages to share your results.

Instructions

Convert to Quarto

The FlukeAndFeather Workflow Organization System™ puts the manuscript Quarto document in the paper/ folder and renders it to the docs/ folder. Why docs/? Because GitHub Pages looks for web pages in the docs/ folder. You’ll see exactly how in the next section..

Make sure you’re on the main branch before proceeding.

  1. Create a Quarto document in paper/ called “paper.qmd”.
  2. Add markdown headers for Background, Data, Methods, and Results.
  3. Migrate your spaghetti code to code chunks under the appropriate headers.
  4. Add markdown text as appropriate to explain your replication.

Don’t render just yet. First we need to set up GitHub Pages.

Set up GitHub Pages

  1. Let’s tell your Quarto document you want the outputs in your docs/ folder. In your paper/ folder, create a file called “_quarto.yml”1. Add the following lines:
project:
  output-dir: ../docs
  1. Render paper.qmd. Verify it created paper.html in docs/.
  2. Navigate to your repo on GitHub. Go to Settings > Code and automation > Pages.
  3. Change your Branch settings to match the image below. Choose the main branch, /docs directory, and hit Save.

  1. Go back to RStudio. Commit and push your changes.
  2. Pop back to your GitHub repo. Switch from Settings to Actions. You should see workflow run(s) deploying your Pages site, similar to the screenshot below.

  1. When the workflow(s) are all green, navigate your browser to [your username].github.io/[reponame]/paper.html. You should see your analysis rendered in all its glory.

Footnotes

  1. Try doing it at the command line!↩︎