Python：ValueError：形状不匹配：形状（3,3）的值数组无法广播到形状（270,3）的索引结果

Question

我正在尝试在 10 折交叉验证中保存每个折叠的多个混淆矩阵实例。 我的代码如下所示：

kf = KFold(n_splits=10, shuffle=True)
conf_matrix = np.zeros([3,3])

for train_index, test_index in kf.split(X):
    X_train, X_test = X.iloc[train_index], X.iloc[test_index]
    y_train, y_test = y.iloc[train_index], y.iloc[test_index]
    
    model.fit(X_train, y_train)
    pred = model.predict(X_test)
    conf_matrix[train_index] = confusion_matrix(y_test, pred)

conf_matrix

但它返回此错误

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-39-8a1e0e2223f7> in <module>()
      9     model.fit(X_train, y_train)
     10     pred = model.predict(X_test)
---> 11     conf_matrix[train_index] = confusion_matrix(y_test, pred)
     12 #     print(confusion_matrix(y_test, pred))
     13 

ValueError: shape mismatch: value array of shape (3,3) could not be broadcast to indexing result of shape (270,3)

这很奇怪，因为如果我运行 print(confusion_matrix(y_test, pred)) 输出是 3x3 矩阵，看起来像这样：

[[4 2 3]
 [4 8 1]
 [1 3 4]]
[[3 6 1]
 [4 5 1]
 [5 1 4]]
[[7 3 4]
 [4 4 1]
 [0 1 6]]
[[3 1 2]
 [2 9 1]
 [3 2 7]]
[[4 4 2]
 [2 8 2]
 [0 2 6]]
[[7 6 1]
 [4 4 2]
 [0 3 3]]
[[4 1 3]
 [2 5 3]
 [3 1 8]]
[[1 4 3]
 [2 5 1]
 [2 3 9]]
[[7 5 2]
 [2 3 3]
 [2 3 3]]
[[4 2 1]
 [2 6 0]
 [3 5 7]]

我哪里做错了？

编辑 2：我已经尝试了 @user1424589 的建议，现在像这样更改我的代码

conf_matrix = np.zeros((10,len(y_test),len(y_test))) #10 because I want it to caontain 10 confusion matrix

但它仍然返回相同的错误，它现在返回此错误

ValueError: shape mismatch: value array of shape (3,3) could not be broadcast to indexing result of shape (270,30,30)

Answer 1

问题是您告诉它要做的是将输出数组分配到具有形状 (3,) 的 conf_matrix 切片中，您需要对 conf_matrix 进行整形，以便它可以接受混淆矩阵的结果（y_test，pred )。