类应包含 y 中的所有有效标签

Question

我想将数据集类别的权重矩阵传递给神经网络。

from sklearn.utils import class_weight
class_weights = class_weight.compute_class_weight('balanced',
                                                 np.unique(y_train),
                                                 y_train)

但是我收到以下错误：

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-93-9452aecf4030> in <module>
      2 class_weights = class_weight.compute_class_weight('balanced',
      3                                                  np.unique(y_train),
----> 4                                                  y_train)

~\AppData\Roaming\Python\Python36\site-packages\sklearn\utils\class_weight.py in compute_class_weight(class_weight, classes, y)
     39 
     40     if set(y) - set(classes):
---> 41         raise ValueError("classes should include all valid labels that can "
     42                          "be in y")
     43     if class_weight is None or len(class_weight) == 0:

ValueError: classes should include all valid labels that can be in y

我不明白，这是我的y_train数据集的一部分：

        grade_A  grade_B  grade_C  grade_D  grade_E  grade_F  grade_G
689526        0        1        0        0        0        0        0
523913        1        0        0        0        0        0        0
266122        0        0        1        0        0        0        0
362552        0        0        0        1        0        0        0

类[A,B,C,D,E,F]包括可以放入 y 中的所有有效标签！

更新

我尝试在数据帧上使用 .values：

from sklearn.utils import class_weight
class_weights = class_weight.compute_class_weight('balanced',
                                                 np.unique(y_train.values),
                                                 y_train.values)

然而它返回了：

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-25-c2342f04abd9> in <module>
      2 class_weights = class_weight.compute_class_weight('balanced',
      3                                                  np.unique(y_train.values),
----> 4                                                  y_train.values)

~\AppData\Roaming\Python\Python36\site-packages\sklearn\utils\class_weight.py in compute_class_weight(class_weight, classes, y)
     38     from ..preprocessing import LabelEncoder
     39 
---> 40     if set(y) - set(classes):
     41         raise ValueError("classes should include all valid labels that can "
     42                          "be in y")

TypeError: unhashable type: 'numpy.ndarray

如果我输入 print(type(y_train)) 我会得到以下答案：

<class 'pandas.core.frame.DataFrame'>

Answer 1

根据文档：

sklearn.utils.class_weight.compute_class_weight(class_weight, classes, y)

classes : ndarray

    Array of the classes occurring in the data, as given by np.unique(y_org) with y_org the original class labels.
y : array-like, shape (n_samples,)

    Array of original class labels per sample;

如果你逃跑type(y_train)，你会得到什么？

您可以将数据帧转换为数组（将 pandas 数据帧转换为 NumPy 数组）：

ytrain = y_train.values