Virtual Environments and pip

When you start building larger applications, you will often want to use third-party packages (libraries written by others) that do not come with Python’s standard library. To prevent different projects from conflicting with each other, we use virtual environments.

Creating a Virtual Environment

A virtual environment is an isolated directory tree that contains a Python installation and additional packages. The venv module is used to create and manage virtual environments. It is part of the standard library, so you can import it directly:

import venv
print(venv.__name__)

venv

A common directory location for a virtual environment is .venv. You can create one by running this command in your terminal:

$ python -m venv .venv

To use the environment, you must activate it. This modifies your shell’s path so that python points to the isolated installation.

$ source .venv/bin/activate

(On Windows, you would run .venv\Scripts\activate instead).

Checking if a Virtual Environment is Active

You can programmatically verify whether your script is running inside a virtual environment. When activated, Python changes sys.prefix to point to the virtual environment directory, while sys.base_prefix continues to point to the system Python installation.

import sys
# We check if we are running inside a virtual environment
# by comparing sys.prefix with sys.base_prefix
in_venv = sys.prefix != getattr(sys, 'base_prefix', sys.prefix)
print(in_venv)

False

(Since you are running this code in the CodeLudo sandbox, no virtual environment is active!)

Installing Packages with pip

Once activated, you use pip to manage packages. pip installs packages from the Python Package Index (PyPI). You can verify pip is available:

import subprocess
res = subprocess.run(['python', '-m', 'pip', '--version'], capture_output=True, text=True)
print('pip' in res.stdout)

True

For example, to install the popular requests library:

$ python -m pip install requests

You can view detailed information about an installed package:

$ python -m pip show requests

Freezing Requirements

To share your project with others, they need to know which packages to install. pip freeze produces a list of installed packages in the format that pip install expects.

$ python -m pip freeze > requirements.txt

Another developer can then clone your project, create their own virtual environment, and install all dependencies in one go:

$ python -m pip install -r requirements.txt