It computes average precision based on the definition of Pascal Competition. It computes the under curve area of precision and recall. Recall follows the normal definition. Precision is a variant. pascal_precision[i] = typical_precision[i:].max()

average_precision_score.py

import numpy as np

def compute_average_precision(test_results, num_true_cases):
    """
    It computes average precision based on the definition of Pascal Competition. It computes the under curve area
    of precision and recall. Recall follows the normal definition. Precision is a variant. 
    pascal_precision[i] = typical_precision[i:].max()
    
    Usage:
        test_results = np.array([True, False, True, False, True, False, False, False, False, True])
        compute_average_precision(test_results, 8) # 0.33
        

        test_results = np.array([1] * 10)
        compute_average_precision(test_results, 8) # 1

        test_results = np.array([0] * 10)
        compute_average_precision(test_results, 8) # 0
        
    Arguments:
        test_results (np.ndarray): test results, i.e [True, False, True, False, False] or [1, 0, 1, 0, 0].
        num_true_cases (int): the number of total true cases.
        
    Returns:
        average_precision score (float)
    """
    true_positive = test_results.cumsum()
    false_positive = (~test_results).cumsum()
    precision = true_positive / (true_positive + false_positive)
    recall = true_positive / num_true_cases
    
    # identical but faster version of new_precision[i] = old_precision[i:].max()
    precision = np.concatenate([[0.0], precision, [0.0]])
    for i in range(len(precision) - 1, 0, -1):
        precision[i - 1] = np.maximum(precision[i - 1], precision[i])
        
    # find the index where the value changes
    recall = np.concatenate([[0.0], recall, [1.0]])
    changing_points = np.where(recall[1:] != recall[:-1])[0]
    
    # compute under curve area
    areas = (recall[changing_points + 1] - recall[changing_points]) * precision[changing_points + 1]
    return areas.sum()