Packages, pip and Virtual Environments

Why this matters

Python’s real power comes from other people’s code: packages from the Python Package Index (PyPI) such as requests or pandas. You install them with pip. Installing into one shared place quickly makes projects clash, so every project gets its own virtual environment.

This lesson needs a terminal on your own computer, so it is marked local-only: the commands cannot run in the browser. The coding exercise (reading and writing a requirements.txt) can be done anywhere.

The idea

Create and use an environment (once per project):

python -m venv .venv                 # create the environment in a folder called .venv
source .venv/bin/activate            # macOS / Linux
.venv\Scripts\Activate.ps1           # Windows PowerShell
python -m pip install requests       # install a package into THIS environment
python -m pip list                   # what is installed
python -m pip freeze > requirements.txt   # save exact versions
python -m pip install -r requirements.txt # recreate it on another machine
deactivate                           # leave the environment

python -m pip makes sure you use the pip that belongs to the python you are running. A requirements.txt has one package per line, optionally with a version specifier:

requests==2.31.0     # exactly this version
flask>=2.0           # this version or newer
numpy                # any version

Never commit the .venv folder itself to Git; commit requirements.txt.

Try it

In a terminal, create a .venv, activate it, and run python -m pip list. It should be almost empty. Install a small package, then run pip list again.

Your turn

Write two functions that work with requirements text:

parse_requirements(text): return a list of (name, specifier) tuples, one per package, in order. Ignore blank lines, full-line comments and # ... comments at the end of a line. Remove spaces from the specifier: "flask >= 2.0" gives ("flask", ">=2.0"), and a bare name gives an empty specifier ("numpy", "").
requirements_text(versions): the opposite. Given a dictionary of name to version (or None for any version), return the file text sorted by name, one line each: "a==2.1\nb==1.0\n".

Recap

pip installs packages from PyPI; use python -m pip.
One virtual environment per project keeps dependencies separate.
requirements.txt records what the project needs; pip freeze writes it.

Common mistakes

Installing without activating the environment, so the package lands in the system Python.
Committing .venv to Git.
ModuleNotFoundError after installing: you installed into a different Python than the one that runs your code.