Question

我是scikit的新手，在尝试使用采样训练集学习者时，我得到了索引越界错误

这里是发生错误的地方

def train_predict(learner, sample_size, X_train, y_train, X_test, y_test): 


    results = {}
    start = time() # Get start time
    learner.fit(X_train[sample_size],y_train[sample_size])
    end = time() # Get end time

    results['train_time'] = end-start

    start = time() # Get start time
    predictions_test = learner.predict(X_test)
    predictions_train = learner.predict(X_train.head(300))
    end = time() # Get end time

    results['pred_time'] = end-start

    results['acc_train'] = accuracy_score(y_train.head(300),predictions_train)

    results['acc_test'] = accuracy_score(y_test,predictions_test)

    results['f_train'] = f_score(y_train.head(300),predictions_train)

    results['f_test'] = f_score(y_test,predictions_test)

    print "{} trained on {} samples.".format(learner.__class__.__name__, sample_size)

    return results

这是主要代码

    clf_A = GaussianNB()
    clf_B = tree.DecisionTreeClassifier()
    clf_C = SVC()

    samples_1 = random.sample(X_train.index,len(X_train)/100)
    samples_10 = random.sample(X_train.index,len(X_train)/10)
    samples_100 = X_train.index

    results = {}
    for clf in [clf_A, clf_B, clf_C]:
        clf_name = clf.__class__.__name__
        results[clf_name] = {}
        for i, samples in enumerate([samples_1, samples_10, samples_100]):
             results[clf_name][i] = \
             train_predict(clf, samples, X_train, y_train, X_test, y_test)


     vs.evaluate(results, accuracy, fscore)

错误在线

---> 21     learner.fit(X_train[sample_size],y_train[sample_size])

它说

IndexError: indices are out-of-bounds

Answer 1

您的错误完全取决于X_train和y_train的样子。

可能适合您情况的常见示例：如果这些是pandas数据框对象，那么修复解决方案可能就像添加.as_matrix（）一样简单：

learner.fit(X_train.as_matrix()[sample_size],y_train.as_matrix()[sample_size])

您可以检查的另一个快速事项是X_train [sample_size]返回的行数和y_train [sample_size]返回的行数是相同的。请注意，这与不与以下评估为true相同，因为X_train [sample_size]可以包含比y_train [sample_size]更多的列：

len(X_train[sample_size]) == len(y_train[sample_size])

在您的问题中提供有关如何构建X_train和y_train的信息或有关其类型和形状的详细信息，可以为您提供更具体的答案。

Answer 2

尝试以下

learner.fit(X_train[**0:sample_size**],y_train[**0:sample_size**])

scikit指数超出范围

2 个答案: