ExpectedCalibrationError

ExpectedCalibrationError

The function calculates the Expected Calibration Error.

Parameters:

• y_data: pd.Series
  True labels.
• probabilities: np.ndarray
  Probabilities of each object in the dataset.
• n_bins: int, default = 20
  Number of bins.

Returns:

• ece: float
  Expected calibration error.

Exceptions:

• TypeError:
  Raised if y_data parameter is not a pandas Series.
  Raised if probabilities parameter is not np.ndarray.
  

• ValueError:
  Raised if n_bins parameter is not a positive integer.
  

Example:

import pandas as pd
import numpy as np
from combat.calibration import ExpectedCalibrationError

# Sample input data
y_data = pd.Series([0, 1, 0, 1, 0])
probabilities = np.array([[0.2, 0.8], [0.6, 0.4], [0.1, 0.9], [0.7, 0.3], [0.3, 0.7]])
n_bins = 20

# Calculate Expected Calibration Error
result = ExpectedCalibrationError(y_data, probabilities, n_bins)
print(result)