Skip to content

Git for Analysts

As data analysis moves towards "Analytics as Code" (e.g., dbt, Python scripts, stored procedures), Version Control using Git is a mandatory skill.

Why Analysts Need Git

  1. Reproducibility: You can always go back to the exact code that generated a specific report.
  2. Collaboration: Multiple analysts can work on the same project without overwriting each other's work (using branches).
  3. Code Review: Peers can review your SQL logic or Python code before it is merged into production.

The Basic Workflow

  1. Clone: Copy the remote repository (from GitHub/GitLab) to your local machine. 
    git clone https://github.com/your-org/your-repo.git
  2. Branch: Create a new workspace for your specific task (never work directly on the main branch). 
    git checkout -b feature/update-churn-query
  3. Edit: Make changes to your SQL, Python, or Markdown files.
  4. Add (Stage): Tell Git which modified files you want to include in the next save. 
    git add my_query.sql
  5. Commit: "Save" the changes with a descriptive message. 
    git commit -m "Updated the churn calculation to exclude test users"
  6. Push: Send your local branch up to the remote repository. 
    git push origin feature/update-churn-query
  7. Pull Request (PR): Go to GitHub/GitLab and open a PR. Ask a peer to review your code. Once approved, it is merged into main.

.gitignore Best Practices

The most important rule for Data Analysts: Never commit raw data to Git!

Git is meant for tracking code, not massive .csv, .parquet, or database dump files. If you commit a 500MB CSV file, it will bloat the repository and potentially leak sensitive PII (Personally Identifiable Information).

Always ensure you have a .gitignore file that ignores data formats: 

# .gitignore
*.csv
*.xlsx
*.parquet
*.sqlite
.env
__pycache__/

Useful Commands

  • git status - Shows which files are modified, staged, or untracked.
  • git log - Shows the history of commits.
  • git pull - Fetches the latest changes from the remote repository to your local machine.

References