pydaptivefiltering 0.7.0

A modern Python implementation of the AdaptiveFiltering toolbox.

pydaptivefiltering

A high-performance Python package for Adaptive Filtering. > Implementations are strictly based on the book Adaptive Filtering: Algorithms and Practical Implementation by Paulo S. R. Diniz.

🛠️ Description

This package provides a modern Python alternative to the MATLAB Adaptive Filtering Toolbox.

🚀 Project Status & Algorithms

The package currently covers the vast majority of classical and advanced linear adaptive algorithms, all validated via pytest with steady-state convergence scenarios.

Current Implementation

The project is currently on its early stages (pre-alpha). The planned order of work for every kind of algorithm is: (Algorithm(1), Examples(2), Notebooks(3)). The following is the planned and current progress for each of the algorithms:

• [11/11] LMS based algorithms
• [2/2] RLS based algorithms
• [5*/5] SetMembership Algorithms
• [4/4] Lattice-based RLS
• [2*/2] Fast Transversal RLS
• [1/1] QR
• [5/5] IIR Filters
• [0/6] Nonlinear Filters
• [0/3] Subband Filters
• [0/4] BlindFilters

PUAP: The SetMembership_Simp_PUAP implementation is currently under technical review and may not exhibit expected convergence in this version.

StabFastRLS: The Fast_TransversalRLS.StabFastRLS implementation is currently under technical review and may exhibit unexpected warnings in this version.

---

💻 Requirements

• Python 3.10+ (Officially tested on 3.12.7)
• NumPy and SciPy

---

Install Instructions

To install the package with pip:

pip install pydaptivefiltering

To install the package with conda:

conda install pydaptivefiltering

---

☕ Examples of uses

A good place to start is with the jupyter notebooks that can be found at the <Examples\Jupyter Notebooks> folder

import pydaptivefiltering as pdf

---

🧪 Technical Validation

The package uses rigorous unit testing to ensure the mathematical integrity of each structure. We currently maintain 57 automated tests covering:

1. Numerical Stability: Floating-point operations for complex structures (Lattice/QR).
2. Convergence Accuracy: Verification of steady-state MSE (Mean Square Error) against system noise levels.
3. Tracking Performance: Capability of algorithms to adapt to abrupt system changes (Step response).

To run tests locally:

pytest

---

☕ Quick Example (Lattice RLS)

import numpy as np
from pydaptivefiltering.LatticeRLS.LRLS_pos import LatticeRLS

# Setup data
n_samples = 2000
x = np.random.randn(n_samples)
d = ... # Your desired signal

# Initialize Filter
model = LatticeRLS(filter_order=10, lambda_factor=0.99, epsilon=0.01)

# Optimization Process
res = model.optimize(x, d)

# Results
print(f"Final Estimation Error: {res['errors'][-1]}")

---

📝 License

This project is under the license found at LICENSE.

---

📖 References

• Diniz, P. S. R. (2020). Adaptive Filtering: Algorithms and Practical Implementation. Springer.
• MATLAB Adaptive Filtering Toolbox.