SheShe
import matplotlib.pyplot as plt
from sklearn.datasets import load_iris
from sheshe import ModalBoundaryClustering
X, y = load_iris(return_X_y=True)
sh = ModalBoundaryClustering().fit(X, y)
sh.plot_classes(X, y)
plt.show()
Smart High-dimensional Edge Segmentation & Hyperboundary Explorer
SheShe converts any probabilistic model into an explorer of its own decision landscape. It follows local maxima of class probability or predicted value to uncover crisp, human‑readable regions that respect the supervised boundary of the problem.
Gallery
The images below showcase example classifications produced by SheShe. They live in the images/ directory; add new files there to extend the gallery and reference them here.
Learn how SheShe explores local maxima and how its tools operate independently in How it works.
Getting started
Install the library and run the basic tests:
pip install sheshe
PYTHONPATH=src pytest tests/test_basic.py::test_import_and_fit -q
See Installation for full instructions.
Performance tips
Performance Tips outlines heuristics and parameter tweaks for large datasets.
Predictors
The project exposes several predictors and utilities. Each section below describes one of them in more detail:
- ModalBoundaryClustering
- SubspaceScout
- ModalScoutEnsemble
- RegionInterpreter
- ShuShu
- CheChe
- InsideForest
- 3D Visualization
Experiments & Benchmarks
Performance comparisons and profiling results are available in Experiments & Benchmarks.
Regression examples
Sample code demonstrating regression training with score_model is provided in Regression examples.
About
SheShe is developed by José Carlos Del Valle. Connect on LinkedIn or explore his portfolio.
Contribute
Improvements are welcome! Fork the repository, install development dependencies, and run the tests:
pip install -e ".[dev]"
PYTHONPATH=src pytest -q
Released under the MIT License.