SEAT – Sutton Entropy Analysis Tool

SEAT is an advanced gravitational wave analysis toolkit designed to test for information retention in black hole mergers through entropy-based signal analysis.

It incorporates classical preprocessing and novel quantum-inspired metrics across multiple entropy models (Shannon, Renyi, Tsallis), enabling high-resolution investigation of potential memory effects, quantum echoes, and topological structures in LIGO strain data.

!!! Currently Undergoing Maintenance: I intend on 'shotgunning' the entropy methodology for renyi and tsallis. For example, for the Tsallis methodology there are more precise q values than simply leaving it at one. There seems to be an ideal range, for example, it could be 1.3 ~ 1.7. I plan on honing in on these ranges and then finding the average value for each point in time across the range.

I imagine for BH mergers, neutron stars, and regular BH data we will find different ideal ranges yet I am unable to clarify that at the moment.!!!

Built from scratch by Matthew R. Sutton — for scientists, by a curious mind.

---

Features

• Full entropy decomposition: Shannon, Renyi, and Tsallis methods
• Monte Carlo event injection system
• Entropy model selection via Bayesian factor analysis
• PSD deviation and ratio testing
• Wavelet entropy visualization
• Mutual information (H1 ↔ L1 correlation)
• Quantum echo detection (across all entropy types)
• Persistent homology and topological tracking (H0, H1 diagrams)
• Soft hair memory estimation with entropy deltas
• Unified event diagnostic output

---

Quickstart

1. Clone the Repository

bash

git clone https://github.com/yourusername/SEAT.git

cd SEAT

2. Pick your event and check your permissions

Ensure that when you before you run the software you have a precise GPS time and event ID ready.

If the GPS time ends in a decimal round to the nearest whole number, since we are taking chunks of the data rather than a specific point.

It is recommended that you use admin priviledges to save the results for each module.

3. Install dependencies

pip install -r requirements.txt

This runs on Python out of terminal or command prompt. Python 3.9+ is recommended.

This was written on Jupyter Notebook and tested on anaconda_prompt.

It is highly recommended to use a similar structure.

4. Run through each module in this order in the terminal/command prompt as some build off of each other

1.) fetching_mechanism.py

2.) preprocess.py

3.) spectrogram.py

4.) entropy_analysis.py

5.) wavelet_analysis.py

6.) monte_carlo.py

7.) bayesian_model.py

8.) cauchy_analysis.py

9.) mutual_information.py

10.) smirnov.py

11.) homology.py

12.) echo.py***

13.) softhairmemory.py

14.) softhairentropy.py

15.) event_diagnostic.py

5. The Quantum echo is optional.

This step is extremely computationally intensive. It took hours on my machine to run just one. This section is explained in detail below.

Measures autocorrelation within post-merger signals to detect potential “echoes”—low-amplitude, delayed repetitions of the original waveform. These echoes are hypothesized signatures of quantum-scale corrections near the event horizon.

• The analysis is run across multiple domains: standard whitened strain data, and entropy-transformed data (Shannon, Renyi, Tsallis).

• Autocorrelation functions are computed and statistically summarized (mean, std) for each channel and method.

• The test looks for non-random structure—persistent post-ringdown signals that cannot be explained by classical noise or detector artifacts.

• Entropy-based echoes offer a novel probe: if echoes persist more clearly in Renyi or Shannon domains than in raw strain, it suggests the information retention may be encoded in statistical structure, not amplitude.

6 Results

After completing the analysis, rename any output files that do not already include the event ID and store them in a dedicated folder to preserve results.

Licensing

SEAT is released under the MIT License for academic and non-commercial use.

Commercial use is prohibited without written permission.

To request a commercial license, contact the author: - Email: [mattsutton9@yahoo.com]

---

TL;DR

• You can use SEAT for academic papers, coursework, and non-profit research
• You can’t package or sell SEAT commercially without permission

Citation

If you use this tool in your research please cite with the following DOI: https://doi.org/10.5281/zenodo.15172345

@misc{seat2025, author = {Matthew R. Sutton}, title = {SEAT - Sutton Entropy Analysis Tool}, year = {2025}, url = {https://github.com/MRaySutt/SuttonEntropyAnalysisTool}, note = {Open-source software for gravitational-wave entropy analysis} }