Python 当只更新一个索引时，numpy zeros 数组为每个值分配 1

Question

以下是我的代码：

amount_features = X.shape[1]

best_features = np.zeros((amount_features,), dtype=int)
best_accuracy = 0
best_accuracy_index = 0

def find_best_features(best_features, best_accuracy):

    for i in range(amount_features):
        trial_features = best_features
        trial_features[i] = 1
        svc = SVC(C = 10, gamma = .1) 
        svc.fit(X_train[:,trial_features==1],y_train)
        y_pred = svc.predict(X_test[:,trial_features==1])
        accuracy = metrics.accuracy_score(y_test,y_pred)
        if (accuracy > best_accuracy):
            best_accuracy = accuracy
            best_accuracy_index = i

    print(best_accuracy_index)
    best_features[best_accuracy_index] = 1

    return best_features, best_accuracy

bf, ba = find_best_features(best_features, best_accuracy)

print(bf, ba)

这是我的输出：

25
[1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1] 0.865853658537

我的预期输出：

25
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0] 0.865853658537

我正在尝试使用提供最高准确度的索引更新 zeros 数组。如您所见，它应该是索引 25，然后我将 25 索引分配给我的数组等于 1。但是，当我打印数组时，它显示每个索引都已更新为 1。

不确定发生了什么事故。感谢您在地球上花费有限的时间来帮助我。

Answer 1

将 trial_features = best_features 更改为 trial_features = numpy.copy(best_features)。 @Michael Butscher 已经给出了更改背后的原因。

Python 当只更新一个索引时，numpy zeros 数组为每个值分配 1

Python numpy zeros array being assigned 1 for every value when only one index is updated

python

numpy

machine-learning

svm

python-3.x