PyWaveSync 0.1.1

WaveSync is a Python library for nuanced, nonlinear, and rapid analysis of vectors and embeddings, tailored for RAG systems.

WaveSync is a Python library crafted for the analysis of vectors and embeddings, specifically tailored for use in Retrieval-Augmented Generation (RAG) systems.

Unlike traditional methods that rely on cosine similarity for embeddings comparison, WaveSync introduces a novel approach by employing time series decomposition and phase analysis. This method enables the library to perform a more nuanced, nonlinear, and rapid analysis of embeddings, allowing for the identification of deeper similarities and differences beyond what cosine similarity can offer.

Installation

Required Libraries

Before installing WaveSync, you have to ensure the following libraries are already installed on your system:

pip install numpy

pip install PyWavelets

Installation WaveSync

pip install PyWaveSync

Usage

For quick start:

from wavesync.wavesync import WaveSync
import numpy as np

vec = np.random.rand(1024)  
vecs = np.random.rand(10, 1024)

ws = WaveSync()

wavesync_scores = ws.compare(vec, vecs)

print(wavesync_scores)

If you want to see how the algorithm works on different types of vectors:

from wavesync.wavesync import WaveSync
import numpy as np

ws = WaveSync()

np.random.seed(42)  
vec = np.random.rand(1024)  
similar_vecs = [vec + np.random.normal(0, 0.01, len(vec)) for _ in range(5)]
dissimilar_vecs = [np.random.rand(len(vec)) for _ in range(5)]
vec_a = np.random.rand(1024)
vec_b = -vec_a  # Coordinate-wise opposite of vec_a

# Test with similar, dissimilar, and opposite vectors
wavesync_similar_scores = ws.compare(vec, similar_vecs)
wavesync_dissimilar_scores = ws.compare(vec, dissimilar_vecs)
wavesync_opposite_scores = ws.compare(vec_a, [vec_b])

print(wavesync_similar_scores, wavesync_dissimilar_scores, wavesync_opposite_scores)

Learn More

For those interested in diving deeper into how the WaveSync algorithm works, detailed explanations and use cases can be found on the following platforms:

• English (Will be available in approximately 3 days): Check out our articles on Medium for an in-depth look at WaveSync.
• Russian (Will be available in approximately 2 days): For Russian speakers, detailed discussions can be found on Habr.

Community and Contact

I'm looking forward to collaborating with anyone interested in improving WaveSync. Your feedback, suggestions, and contributions are always welcome.

How to Contribute

• Contributions: If you'd like to contribute, start by forking the repository on GitHub. Then, create a new branch for your feature or bug fix, make your changes, and test them. When you're ready, submit a pull request with a detailed description of your work.

• Feedback and Discussions: For comments, questions, or suggestions, please use GitHub Issues. It's a great way to provide feedback or start a conversation about the library.

Direct Contact

• If you have specific inquiries or ideas you'd prefer to discuss directly, you can reach out to me via email at [e-mail]. I'm always open to hearing from users and potential collaborators.

Your involvement is crucial for making WaveSync even better.

License

This project is licensed under the MIT License - see the LICENSE file for details.