Disk Space¶

Professional Python projects use isolated environments. Each project installs its own packages into a .venv folder. This is intentional to prevent version conflicts between projects. The tradeoff is disk space.

How Much Space to Expect¶

Project Type	Typical `.venv` Size	Typical Package Count
Basic Python + dev tools + zensical docs	~100 MB	~47
+ Data files (openpyxl, csv)	~100 MB	~49
+ SQL (duckdb, pandas)	~200 MB	~51
+ Notebooks (jupyter, pandas, matplotlib, seaborn)	~350–400 MB	~80–85
+ Machine learning (scikit-learn added)	~400–450 MB	~88–92

These sizes reflect installed packages only. Project source code and data files are small by comparison.

One Active Project at a Time¶

The recommended workflow is to have one project open at a time. Each project lives in its own folder under Repos/ with its own .venv. When you finish a project and push to GitHub, you can delete the local project repository folder entirely.

Peak disk usage following this practice is the size of one project, or about 100–450 MB depending on the project.

Deleting a Project to Reclaim Space¶

Once work is in GitHub (verify first!), it can be deleted from the machine.

Everything needed to restore the project is in GitHub. Project source code, data files, notebooks, and dependency list are all committed. The .venv in not in Git and should not be, because uv sync can rebuild it from uv.lock typically in less than a minute.

Restoring a Project¶

To pick up where you left off, use git clone to copy your GitHub repo to your machine. And follow the steps to recreate your environment using uv.

uv python pin 3.14
uv sync --extra dev --extra docs --upgrade

Safe to Delete¶

Path	Safe to delete?	Notes
`.venv/`	Yes	Rebuilt by `uv sync`
`site/`	Yes	Rebuilt by `uv run python -m zensical build`
`.coverage`	Yes	Rebuilt by `uv run python -m pytest`
`__pycache__/`	Yes	Rebuilt automatically on next run

Keep These (Ensure They are Safe in GitHub)¶

Path	Safe to delete?	Notes
`project.log`	Commit first	Generated output, commit before deleting
`data/processed/`	Commit first	Generated output, commit before deleting
`uv.lock`	Commit first	Commit this to ensure reproducible installs
`pyproject.toml`	Commit first	Commit this as it defines the project
`src/`	Commit first	Your code lives here
`notebooks/`	Commit first	Your analysis lives here

Dependabot and Long-Term Maintenance¶

If you keep a project repo active on GitHub, Dependabot may open pull requests to update outdated packages. This is a security and maintenance feature, not an error.

For active course repos: you do not need to act on Dependabot PRs.

After the course: if you want to keep a repo as a portfolio piece, merging Dependabot PRs occasionally keeps dependencies current and signals active maintenance. If you let a repo go inactive, Dependabot will eventually stop opening PRs automatically.

If you get a Dependabot PR: the safest action is to merge it, then run:

git pull
uv sync --extra dev --extra docs --upgrade
uvx pre-commit run --all-files

Verify everything still runs before pushing.