How alejandrosantorum/scikit-rmt works
scikit-rmt is a Python library for random-matrix sampling and spectral analysis, with a separate RMT-based image-denoising API. Users sample an ensemble, inspect eigenvalues or histograms, and compare results with spectral laws; image-stack users can denoise with MP-PCA and evaluate image quality. The graph includes the documented ensemble families and laws, plus the sampled Wishart-to-denoiser dependency. Most ensemble implementations were not sampled, so uncertain internal call relationships are omitted.
Main entry points: wishart_ensemble.py Wishart ensemble · base_ensemble.py Ensemble base · mp_pca.py MP-PCA denoiser · spectral_law.py Spectral distributions
Components 15
Random ensembles
- Gaussian ensemble
skrmt/ensemble/gaussian_ensemble.py - Wishart ensemble
skrmt/ensemble/wishart_ensemble.py - Manova ensemble
skrmt/ensemble/manova_ensemble.py - Circular ensemble
skrmt/ensemble/circular_ensemble.py
Spectral laws
- Spectral distributions
skrmt/ensemble/spectral_law.py - Tracy-Widom approximation
skrmt/ensemble/tracy_widom_approximator.py
Image denoising
- MP-PCA denoiser
skrmt/denoise/mp_pca.py - Image quality metrics
skrmt/denoise/metrics.py
Shared internals
- Ensemble base
skrmt/ensemble/base_ensemble.py - Ensemble utilities
skrmt/ensemble/utils.py - Tridiagonal tools
skrmt/ensemble/tridiagonal_utils.py
Other components
- Python user
- Noisy image stack
- Denoised stack
- Quality scores
Connections 18
- Python user → Gaussian ensemble: sample and analyze
- Python user → Wishart ensemble: sample and analyze
- Python user → Manova ensemble: sample and analyze
- Python user → Circular ensemble: sample and analyze
- Gaussian ensemble → Ensemble base: uses base methods
- Wishart ensemble → Ensemble base: uses base methods
- Manova ensemble → Ensemble base: uses base methods
- Circular ensemble → Ensemble base: uses base methods
- Wishart ensemble → Spectral distributions: configures MP law
- Python user → Spectral distributions: evaluate law
- Spectral distributions → Tracy-Widom approximation: uses approximation
- Python user → Noisy image stack: provides input
- Noisy image stack → MP-PCA denoiser: transforms
- MP-PCA denoiser → Wishart ensemble: gets MP edge
- MP-PCA denoiser → Denoised stack: returns
- Python user → Image quality metrics: evaluate output
- Denoised stack → Image quality metrics: compared with reference
- Image quality metrics → Quality scores: returns scores