第12章 监督学习方法总结 - 交叉验证实现

import numpy as np

def k_fold_split(X, y, k=5, shuffle=True, random_state=None):
    n_samples = len(X)
    indices = np.arange(n_samples)
    
    if shuffle:
        if random_state is not None:
            np.random.seed(random_state)
        np.random.shuffle(indices)
    
    fold_sizes = np.full(k, n_samples // k)
    fold_sizes[:n_samples % k] += 1
    
    folds = []
    current = 0
    for fold_size in fold_sizes:
        test_indices = indices[current:current + fold_size]
        train_indices = np.concatenate([indices[:current], indices[current + fold_size:]])
        folds.append((train_indices, test_indices))
        current += fold_size
    
    return folds

def cross_val_score(model, X, y, k=5, shuffle=True, random_state=None):
    folds = k_fold_split(X, y, k, shuffle, random_state)
    scores = []
    
    for train_idx, test_idx in folds:
        X_train, X_test = X[train_idx], X[test_idx]
        y_train, y_test = y[train_idx], y[test_idx]
        
        model.fit(X_train, y_train)
        score = model.score(X_test, y_test)
        scores.append(score)
    
    return np.array(scores)