pip and Packages Beyond the Standard Library

The standard library is huge, but the true power of Python is PyPI — the Python Package Index — with over 500,000 free packages. One command installs almost any capability: web frameworks, AI libraries, PDF tools, you name it.

pip — your package manager

pip is Python's installer (it ships with Python). The essential commands:

code

pip install package-name          # install
pip install package-name==2.1.0   # install a SPECIFIC version
pip install --upgrade package-name # update
pip uninstall package-name        # remove
pip list                          # what's installed
pip show package-name             # details about one package

Install something real — the requests library (next lesson's star):

code

pip install requests

Then import it like the standard library:

code

import requests
print(requests.__version__)

One command, one import — that's the entire deal. This is why Python dominates: whatever you're building, someone probably published the hard parts already.

Always inside a virtual environment

Module 2 taught venvs — this is where they matter most:

code

# the ritual for every new project
python -m venv .venv
source .venv/bin/activate        # Windows: .venv\Scripts\activate
pip install requests pandas

Without a venv, every package lands in one global pile — version conflicts between projects are inevitable. With a venv, each project owns its packages. Install globally = future chaos. Install in a venv = future calm.

requirements.txt — your project's recipe

Freeze what you've installed so anyone (including future you, including deployment servers) can recreate the exact environment:

code

pip freeze > requirements.txt

The file looks like:

code

requests==2.31.0
streamlit==1.41.1
pandas==2.2.3

Recreating the environment anywhere:

code

pip install -r requirements.txt

This two-file pair (code + requirements.txt) is what makes projects portable — and it's exactly what the Streamlit Cloud deployment in your other course requires.

Finding packages — how to choose

When you need a capability, search PyPI.org or ask: "python how to [task]". With thousands of options, pick by:

Downloads — PyPI shows monthly numbers; high = battle-tested

Recent updates — abandoned packages rot as Python evolves

Documentation — a package with good docs saves hours

The famous ones you'll meet on your journey:

Package	Superpower
`requests`	HTTP/API calls
`pandas`	data analysis
`streamlit`	web apps from Python
`numpy`	fast number crunching
`matplotlib`	charts
`beautifulsoup4`	web scraping
`openai` / `mistralai`	AI APIs

Common Errors & Fixes

`pip: command not found` — try python -m pip install ... (runs pip through your Python, always works).

Installed but `ImportError` — installed into a different Python than the one running. Check: which python and which pip point to the same venv.

`Permission denied` — installing globally without rights. The real fix: use a venv (never sudo pip).

Version conflicts — two projects need different versions of the same package. The entire reason venvs exist — separate them.

✅ Checkpoint

Install requests version 2.31.0 exactly? *(pip install requests==2.31.0)*

What file makes a project reproducible, and what creates it? *(requirements.txt — pip freeze > requirements.txt)*

Why install inside a venv? *(Per-project isolation — no version conflicts)*

pip: command not found — what always works? *(python -m pip install ...)*

Next: requests — your code's passport to the internet.