Python virtual environments¶

Venv¶

The purpose of venv is to create isolated Python environments for individual projects. This allows each project to have its own dependencies and package versions without interfering with other projects or the system-wide Python installation. It ensures cleaner organization, avoids version conflicts, and enables easy reproduction of the environment using tools like requirements.txt. This is especially useful when collaborating with others or deploying code across different systems.

Create a Project Directory
Terminal
```
mkdir my_project
cd my_project
```
Create a Virtual Environment
Terminal
```
python -m venv venv
```
This creates a venv/ folder containing the isolated Python environment.

You can name it anything (e.g., .venv is common too).
Activate the Virtual Environment
Terminal
```
source venv/bin/activate
```
Install Dependencies

Use pip to install libraries:
Terminal
```
pip install requests flask
```
Freeze Dependencies Save your dependencies:
Terminal
```
pip freeze > requirements.txt
```
Create a .gitignore (if using git)

Prevent venv/ from being tracked:
Terminal
```
echo venv/ __pycache__/ *.pyc >> .gitignore
```
Run Your Python App You can now run your scripts:
Terminal
```
python main.py
```

Reproducing the Environment (for others) To recreate the environment:

Terminal

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Comparison of venv vs conda¶

Feature	`venv` (Standard Python)	`conda` (Anaconda/Miniconda)
Purpose	Create isolated Python environments	Manage environments and packages (Python + non-Python)
Included With	Standard in Python 3.3+	Requires installing Anaconda or Miniconda
Language Support	Python only	Supports multiple languages (Python, R, Julia, etc.)
Package Manager	Uses `pip`	Uses `conda` (can also use `pip` inside conda)
Speed of Package Resolution	Fast for pure Python; slower for complex dependencies	Faster for scientific packages due to prebuilt binaries
System Packages	Installs from PyPI	Installs from conda channels (e.g., conda-forge)
Non-Python Dependencies	Manual (e.g., via apt/brew)	Built-in (e.g., OpenCV, HDF5, BLAS)
Binary/Compiled Packages	Not handled natively; relies on wheels from PyPI	Conda packages are often precompiled
Environment Reproducibility	Via `requirements.txt`	Via `environment.yml`
Cross-Platform Consistency	Less consistent (due to pip building from source)	More consistent across OSes
Footprint	Lightweight (just Python + pip)	Heavier (especially full Anaconda distribution)