How to Use `venv` in Python 3.12

This story simply explains how to use venv in Python 3.12 (precisely speaking, Python 3.7 or later).

Environment

• Python 3.12.1

Premises

venv is a module to create isolated environments for managing project-specific packages. In other words, it enables us to create separate pip environments.

Actions

0: Go to the directory where you want to set up venv

cd <path/to/dir>

1: Create a new environment

.venv is a directory to store everything about the virtual environment. The name can be anything, but I personally prefer using .venv.

python -m venv .venv

If you use git, you may want to add this line, .venv/, to .gitignore.

2: Activate the environment

Mac OS and Windows have different commands.

# For Mac OS
source .venv/bin/activate

# For Windows PowerShell
.venv\Scripts\Activate.ps1

If the environment is successfully activated, every line of the Terminal shows (.venv), or the name you defined at Step 1.

3: Use pip to install packages

Use these pip commands to install packages. Usually we use requirements.txt to install packages required by each project.

# Install packages written in requirements.txt
pip install -r requirements.txt

# Create requirements.txt based on the current environment 
pip freeze > requirements.txt

If you want to install one by one, just write a package name after pip install. Note that <package name> can be <package name>==<version>, such as numpy==2.1.3, if you need a specific version.

# Install a new package
pip install <package name>

# Upgrade a package
pip install -U <package name>

# Forcely reinstall a package (You may want to do this when there is something wrong.)
pip install --force-reinstall <package name>

Additionally, pip has these helpful functions.

# Show all installed packages
pip list

# Show detailed info about a specific installed package
pip show <package name>

#### Example: the result of `pip show ipython` ####
# Name: ipython
# Version: 8.15.0
# Summary: IPython: Productive Interactive Computing
# Home-page: https://ipython.org
# Author: The IPython Development Team
# Author-email: ipython-dev@python.org
# License: BSD-3-Clause
# Location: /opt/homebrew/lib/python3.12/site-packages
# Requires: appnope, backcall, decorator, jedi, matplotlib-inline, pexpect, pickleshare, prompt-toolkit, pygments, stack-data, traitlets
# Required-by: ipykernel

4: Deactivate the environment

Very simple.

deactivate

5: Remove the environment

This is also straightforward. Remove the .venv directory.

# Assume the current directory has `.venv`
# For Mac OS
rm -rf .venv  

# For Windows PowerShell
Remove-Item -Path ".venv" -Recurse -Force

Actions for VSCode & Jupyter Notebook

1: Open a New Jupyter Notebook

Open a new .ipynb file and click “Select Kernel” on the right.

2: Click “Python Environments…”

3: Choose a Python Environment

The path to the venv environment should be shown like .venv/bin/python.

References

venv - Creation of virtual environments