Dependency Management

Pytron provides a zero-config dependency management workflow. It automatically handles virtual environments so you don't have to worry about conflicting packages or system-wide installations.

The Pytron Venv

When you initialize a project or run an install command, Pytron creates a specialized virtual environment (usually in a hidden .pytron_env or similar internal directory). This environment is used exclusively by your app during development and packaging.

Commands

Installing Packages

Use the pytron install command to add packages. This automatically detects your OS and installs the correct binaries.

# Install a single package
pytron install numpy

# Install multiple packages
pytron install pandas requests toggl-python

The requirements.json

Instead of a standard requirements.txt, Pytron uses a modern requirements.json file to track dependencies. This allows for better metadata and platform-specific targeting in the future.

{
  "dependencies": [
    "pytron-kit",
    "numpy",
    "requests"
  ]
}

Running pytron install without any arguments will sync your virtual environment with the requirements.json.

Uninstalling

pytron uninstall numpy

This removes the package from the internal venv and updates the requirements.json manifest automatically.

Complex Dependencies

For heavy libraries like PyTorch, TensorFlow, or llama-cpp-python, use the force-package flag in your settings.json to ensure the compiler (Nuitka or Shield) includes the necessary shared libraries (.dll/.so) in the final bundle.