venv is a tool that creates isolated Python environments, allowing you to manage different sets of packages for different projects. Each environment acts as a sandbox, avoiding conflicts with the system-wide Python installation.
-
Isolation: Packages in a virtual environment won't affect the system Python or other environments.
-
Package Management: Different package versions can be installed for each project.
-
Dependency Control: Ensures project stability by defining specific package versions.
-
Activation/Deactivation: Easily switch between environments for different projects.
-
Cross-platform Compatibility: Works on various operating systems.
-
Create a virtual environment:
python -m venv .venv -
Activate the environment:
- Windows:
.venv\Scripts\activate - macOS/Linux:
source .venv/bin/activate
- Windows:
-
Deactivate:
deactivate
venv: Built-in, lightweight, Python-specific, usespip, limited handling of native dependencies.conda: External, cross-platform, multi-language, usesconda, excels with native dependencies, popular in data science.
-
venv: You can create a virtual environment with venv, install common packages usingpip, and then reuse that environment for multiple projects. This approach allows you to maintain consistency between projects that need the same set of packages. But usage will be harder - you will need to know where activation script is located and run it manually. -
conda: Conda environments are well-suited for creating global reusable environments because they support multi-language packages and can handle complex dependencies more effectively. You can create an environment with the necessary packages using conda, and then use that environment as a base or template for different projects.