Uniform Error Estimator (UEE) is a Python implementation of minimax (Chebyshev / L∞) regression:

[ \min_�eta \max_t |y_t - X_t �eta| ]

• Pypi package: 'uee'
• Source repository: 'uniform-error-estimator'

The package provides:

• a clean estimator API (UEE)
• rich fit diagnostics (UEEResults)
• bootstrap inference
• optional intercept handling
• active-set and dual-weight diagnostics

pip install uee

import numpy as np
from uee import UEE

X = np.array(\\\[\\\[1.0], \\\[2.0], \\\[3.0], \\\[4.0]])
y = np.array(\\\[2.1, 4.0, 5.9, 8.2])

model = UEE(fit\\\_intercept=True)
res = model.fit(X, y)

print(res.intercept\\\_)
print(res.coef\\\_)
print(res.max\\\_abs\\\_error)
print(res.active\\\_set)

This repository/package is the maintained, packaged version of UEE. If you previously exposed a raw script or notebook implementation, update that legacy location to point users here.

MIT. See LICENSE.