ROC_CURVE- IndexError：数组的索引过多

Question

分类，当我输入具有测试标签和测试概率的numpy数组时，它将引发以下错误

dataset = read_csv('C:/.../dataset/KDDREAL.csv')
dataset = dataset.values
X = dataset[:, :-1]
Y = dataset[:, -1]

encoder = LabelEncoder().fit(Y)
encoded_Y = encoder.transform(Y)


X_train, X_test, Y_train, Y_test = train_test_split(X , encoded_Y , test_size=0.3, random_state=0)

model_svm = svm.SVC()
model_svm.fit(X_train, Y_train) 
results_svm = model_svm.predict(X_test)

fpr2 = dict()
tpr2 = dict()
roc_auc2 = dict()
for i in range(num_classes):
    fpr2[i], tpr2[i], _ = roc_curve(Y_test , results_svm[:, i])
    roc_auc2[i] = auc(fpr2[i], tpr2[i])
    # Compute micro-average ROC curve and ROC area
fpr2, tpr2, _ = roc_curve(y_test.ravel(), PGD20_X_test_trm.ravel())
roc_auc2 = auc(fpr2, tpr2)

fpr2[i], tpr2[i], _ = roc_curve(Y_test , results_svm[:, i])

IndexError: too many indices for array

Answer 1

对于下面给出的代码行-

for i in range(num_classes):
       fpr2[i], tpr2[i], _ = roc_curve(Y_test , results_svm[:, i])

尝试将其替换为-

for i in range(n_classes):
      fpr[i], tpr[i], _ = roc_curve(Y_test[:, i], results_svm[:, i])

如果您要复制此行，请检查缩进。