Skip to main content
Angreal Documentation
Toggle Dark/Light/Auto mode Toggle Dark/Light/Auto mode Toggle Dark/Light/Auto mode Back to homepage
  1. Angreal Documentation
  2. /
  3. How-to Guides
  4. /
  5. Work with Virtual Environments
Edit page

Work with Virtual Environments

Work with Virtual Environments

Angreal provides virtual environment management using UV, a fast Python package installer and resolver.

Overview

The virtual environment integration provides:

  • Create and manage virtual environments
  • Install packages and requirements
  • Ensure specific Python versions
  • Automatic environment activation for tasks
  • Programmatic activation within Python processes

Basic Usage 

from angreal.integrations.venv import VirtualEnv


# Create a new environment
venv = VirtualEnv("myproject-env")

# Install packages
venv.install(["django", "requests"])

# Install from requirements file
venv.install("requirements.txt")

Activating Virtual Environments

Manual Activation 

from angreal.integrations.venv import VirtualEnv

# Create and activate a virtual environment
venv = VirtualEnv("data-env", now=True)
venv.install(["numpy", "pandas"])

# Activate the environment
venv.activate()

# Now you can import the installed packages
import numpy as np
import pandas as pd

# Remember to deactivate when done
venv.deactivate()

Using Context Manager 

from angreal.integrations.venv import VirtualEnv

# Automatic activation/deactivation with context manager
with VirtualEnv("analysis-env", now=True) as venv:
    venv.install("scikit-learn")

    # Environment is activated here
    from sklearn import datasets
    iris = datasets.load_iris()

# Environment is automatically deactivated here

Using with Tasks 

from angreal.integrations.venv import venv_required
import angreal

@angreal.command(name="analyze", about="Run data analysis")
@venv_required(".venv", requirements=["pandas", "numpy"])
def analyze_data():
    """This function runs in an isolated environment with packages available."""
    import pandas as pd
    import numpy as np
    # Your code here - the virtual environment is activated

Python Version Management 

from angreal.integrations.venv import VirtualEnv


# Create environment with specific Python version
venv = VirtualEnv("py312-env", python="3.12")

# List available Python versions
pythons = VirtualEnv.discover_available_pythons()
for version, path in pythons:
    print(f"{version}: {path}")

Next Steps