Leymosun: High-Entropy Statistical Research Toolkit
12 days ago
- #randomness-research
- #high-entropy
- #matrix-ensembles
- Leymosun is a High-Entropy Randomness Research Toolkit for randomness-based research.
- Features High-Entropy Random Number Generation (HE-RNG) with non-deterministic seeds for improved simulation quality.
- Includes tools for generating random matrices in canonical and mixed ensembles, spectral analysis, and analytic distributions like Wigner's semicircle law.
- Emphasizes scientific correctness over speed, with potential for faster implementations of proven methods.
- Provides HE-RNG utilities with NumPy backend, including various distributions (Binomial, Uniform, Normal).
- Supports generation of Gaussian and Mixed Gaussian ensembles, spectra generation, and extraction of offdiagonal elements.
- Includes data manager utilities and lecture notes on randomization concepts, Wigner's semicircle law, and NIST benchmark tests.
- Used in publications like 'Empirical deviations of semicircle law in mixed-matrix ensembles' by M. Suzen (2021).
- Released under GPLv3 for code and CC BY 4.0 for documentation.
- Project details include release history, download options, and file metadata for version 0.1.55.