Skip to content

Latest commit

 

History

History
175 lines (116 loc) · 3.89 KB

INSTALL.md

File metadata and controls

175 lines (116 loc) · 3.89 KB

Installation Guide for PyBDR

PyBDR is a set-boundary based reachability analysis toolbox for Python.

Prerequisites

  • Conda or Miniconda (Download here)
  • Git (for cloning the repository)

Quick Installation (3 Steps)

# Step 1: Create conda environment with all dependencies
conda env create -f environment.yml

# Step 2: Activate the environment
conda activate pybdr

# Step 3: Install PyBDR
pip install -e .

That's it! The environment.yml automatically installs Python 3.8+, all dependencies, and the required cddlib C library.

Alternative: Install in Existing Conda Environment

If you already have a conda environment:

# Activate your environment
conda activate your_env_name

# Install cddlib C library
conda install -c conda-forge cddlib

# Install PyBDR (this will install all Python dependencies including codac pre-release)
pip install -e .

For Developers

# Clone and setup
git clone https://github.com/ASAG-ISCAS/PyBDR.git
cd PyBDR
conda env create -f environment.yml
conda activate pybdr

# Install with development dependencies
pip install -e ".[dev]"

# Run tests
pytest

Verifying Installation

After installation, verify that PyBDR is correctly installed:

import pybdr
from pybdr.geometry import Interval, Zonotope
from pybdr.algorithm import ASB2008CDC

# Create a simple interval
interval = Interval([0, 1], [1, 2])
print(f"Interval created: {interval}")

print("PyBDR successfully installed!")

Troubleshooting

Issue: pycddlib build fails with "cddlib/setoper.h not found"

Cause: The cddlib C library is not installed.

Solution: Make sure you've installed cddlib via conda:

conda install -c conda-forge cddlib

Issue: CVXPY installation fails

Solution: CVXPY is included in environment.yml. If you're installing manually, use conda:

conda install -c conda-forge cvxpy

Issue: "No module named 'pybdr'" after installation

Solution: Make sure you've activated the correct conda environment:

conda activate pybdr

Issue: Codac installation or "No module named 'codac'"

Cause: Codac v2 is only available as a pre-release version.

Solution: Codac pre-release is automatically included when you run pip install -e .. If you need to install it manually:

pip install "codac>=2.0.0.dev20"

Or reinstall PyBDR:

pip install -e . --force-reinstall

Package Structure

After installation, PyBDR provides the following modules:

pybdr/
├── algorithm/          # Reachability analysis algorithms
├── dynamic_system/     # System models (continuous, discrete, hybrid)
├── geometry/           # Geometric representations (intervals, zonotopes, polytopes)
├── model/              # Pre-defined system models
├── util/               # Utility functions
└── misc/               # Miscellaneous utilities

Updating PyBDR

If you installed in development mode (pip install -e .), simply pull the latest changes:

cd PyBDR
git pull origin master

Uninstalling

To remove PyBDR:

pip uninstall pybdr

To remove the entire conda environment:

conda deactivate
conda env remove -n pybdr

Support

For issues and questions:

Dependencies

PyBDR depends on:

  • numpy (≥1.20.0): Numerical computations
  • scipy (≥1.7.0): Scientific computing
  • pypoman (≥0.5.4): Polytope manipulation
  • cvxpy (≥1.1.0): Convex optimization
  • matplotlib (≥3.3.0): Visualization

System dependencies (automatically installed via conda):

  • cddlib: C library for polytope computations (required by pypoman)

All dependencies are automatically installed when using the environment.yml file with conda.