Git and GitHub for Analysts

Step 1 — Install Git and set your identity (once)

1. Install Git from git-scm.com/downloads (all defaults are fine).
2. Optionally install the GitHub CLI from cli.github.com. It makes creating repos and signing in much easier.
3. Open a terminal (Git Bash on Windows, Terminal on Mac) and tell Git who you are. This name and email get stamped on every save:

git config --global user.name "Your Name"
git config --global user.email "you@example.com"

If you installed the GitHub CLI, sign in to GitHub once with gh auth login and follow the prompts. This is what lets you push without pasting a password every time.

Step 2 — The everyday loop (this is 90% of Git)

Every time you finish a chunk of work, you do the same three-beat move: stage, save, send. In a terminal opened inside your project folder:

git status                 # see what changed (read this first, always)
git add .                  # stage every changed file for the next save
git commit -m "Add churn analysis notebook"   # save, with a short note
git push                   # send it up to GitHub

That is the whole daily rhythm. status shows you what is about to be saved, add chooses the files, commit saves a snapshot with a message, and push uploads it. Do this often, with clear messages. Think of each commit as a labeled save point you can return to.

Step 3 — Start a brand-new project and put it on GitHub

You have a folder with your work in it and you want it on GitHub. From inside that folder:

git init -b main           # start tracking this folder, on a branch called main
git add .                  # stage everything
git commit -m "First commit"   # your first save point

Now create the online home for it. The easy way, with the GitHub CLI, does it in one line and pushes at the same time:

gh repo create my-project --public --source=. --remote=origin --push

No CLI? Create an empty repository on github.com/new (do not add a README, keep it empty), then copy the two commands GitHub shows you under "push an existing repository." They look like this:

git remote add origin https://github.com/yourname/my-project.git
git push -u origin main

Step 4 — Publish it as a live website (GitHub Pages)

If your project has an index.html, GitHub can host it as a real, shareable web page for free. On the repo's page: Settings → Pages → Source: Deploy from a branch → Branch: main / (root) → Save.

Wait one to two minutes for the first build, then visit:

https://yourname.github.io/my-project/

Step 5 — What NOT to commit

Git is for your code and small files, not giant datasets. GitHub rejects files over 100 MB, and committing a 2 GB CSV or a database file is a mistake either way. You exclude files by listing them in a .gitignore file in your project root:

# .gitignore
*.db            # SQLite databases
*.csv           # large raw data
data/raw/       # a whole folder of source files
node_modules/   # dependencies

The rule of thumb: commit the things needed to rebuild the project (code, a small sample file, a README), not the heavy data itself. Anyone can re-download or regenerate the big files.