scorio Documentation

scorio implements the Bayes@N framework for Bayesian performance evaluation with uncertainty quantification.

Installation

Install from PyPI:

pip install scorio

Quick Start

import numpy as np
from scorio import eval

# Binary outcomes: M=2 questions, N=5 trials
R = np.array([[0, 1, 1, 0, 1],
              [1, 1, 0, 1, 1]])

# Rubric weights for binary outcomes
w = np.array([0.0, 1.0])

# Bayesian evaluation
mu, sigma = eval.bayes(R, w)
print(f"μ = {mu:.4f}, σ = {sigma:.4f}")

# Pass@k metrics
pass_k = eval.pass_at_k(R, k=2)
print(f"Pass@2 = {pass_k:.4f}")

Contents

Additional Information

Indices and tables