kalbee 0.1.0

A clean, modular Python implementation of Kalman Filters and estimation algorithms.

kalbee 🐝

kalbee is a clean, modular Python implementation of Kalman Filters and related estimation algorithms. Designed for simplicity and performance, it provides a standard interface for state estimation in various applications.

✨ Features

• Standard Kalman Filter (KF): Linear estimation for dynamic systems.
• Extended Kalman Filter (EKF): Support for non-linear systems using Jacobians.
• Alpha-Beta-Gamma Filter: Lightweight filtering for kinematic tracking.
• AutoFilter: Factory pattern for easy filter instantiation.
• Modular Design: Easy to extend with new filter types.
• NumPy Integration: Optimized for numerical computations.
• Robustness: Uses Joseph form for numerically stable covariance updates.

🚀 Installation

You can install kalbee directly from the source:

git clone https://github.com/MinLee0210/kalbee.git
cd kalbee
pip install .

Or using uv (recommended):

uv pip install -e .

🛠️ Quick Start

1. Standard Kalman Filter

import numpy as np
from kalbee.modules.filters import KalmanFilter

# Define matrices
state = np.zeros((2, 1))  # [position, velocity]
cov = np.eye(2)
F = np.array([[1, 1], [0, 1]])  # Constant velocity model (dt=1)
Q = np.eye(2) * 0.01
H = np.array([[1, 0]])          # We measure position
R = np.array([[0.1]])

# Initialize
kf = KalmanFilter(state, cov, F, Q, H, R)

# Predict & Update
kf.predict()
kf.update(np.array([[1.2]]))

print(f"Estimated State:\n{kf.x}")

2. AutoFilter Factory

Easily switch between filters using the AutoFilter factory:

from kalbee.modules.filters import AutoFilter

# Create an EKF
ekf = AutoFilter.from_filter(
    state, cov, Q, R,
    mode='ekf',
    transition_function=my_f,
    measurement_function=my_h
)

📚 Documentation

Detailed documentation and examples can be found in the docs/ directory.

• Getting Started
• Filtering Logic

🧪 Testing

Run these commands to test the library:

uv run pytest tests/

📄 License

This project is licensed under the Apache License 2.0.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. Check TODO.md for ideas.