samronsin

import numpy as np
from sklearn.metrics import calibration_loss

def sample_calibrated(N):
    xs = []
    ys = []
    for _ in range(N):
        x = np.random.rand()
        xs.append(x)
        y = 1 if np.random.rand() < x else 0