在 Kaggle 内核上使用 train_test_Split 方法时显示错误。 帮助表示赞赏

Question

任何人都可以解释什么是错误以及如何避免？ 相同的代码实际上适用于我的 Jupyter 和 Spider 环境。 我也不明白为什么这不适用于 kaggle 内核。

下面是我的代码。

# Prepare X and y objects for sikitlearn
X = df.iloc[:, 1:5].values
y = df.iloc[:,-1].values

# Splitting the Dataset into the training set and the test set
X_train, X_test, y_train, y_test = train_test_split(X,y, test_size = 0.3, random_state =4)

错误详情：

        ---------------------------------------------------------------------------
    TypeError                                 Traceback (most recent call last)

<ipython-input-49-5647b5a0596a> in <module>()
      1 # Splitting the Dataset into the training set and the test set
      2 from sklearn.model_selection import train_test_split
----> 3 X_train, X_test, y_train, y_test = train_test_split(X,y, test_size = 0.3, random_state =4)

/opt/conda/lib/python3.6/site-packages/sklearn/model_selection/_split.py in train_test_split(*arrays, **options)
   2100 
   2101     return list(chain.from_iterable((safe_indexing(a, train),
-> 2102                                      safe_indexing(a, test)) for a in arrays))
   2103 
   2104 

/opt/conda/lib/python3.6/site-packages/sklearn/model_selection/_split.py in <genexpr>(.0)
   2100 
   2101     return list(chain.from_iterable((safe_indexing(a, train),
-> 2102                                      safe_indexing(a, test)) for a in arrays))
   2103 
   2104 

/opt/conda/lib/python3.6/site-packages/sklearn/utils/__init__.py in safe_indexing(X, indices)
    183                                    indices.dtype.kind == 'i'):
    184             # This is often substantially faster than X[indices]
--> 185             return X.take(indices, axis=0)
    186         else:
    187             return X[indices]

TypeError: take_nd() got an unexpected keyword argument 'axis'

Answer 1

可能是因为您的目标变量的类型不是分类的，您必须将其更改为标签。

Answer 2

这是一个尚未解决的错误。

为了使代码工作，您的响应变量不应是分类的。

X_train, X_test, y_train, y_test = train_test_split(data_frame, y_was_categorical.astype(str), stratify=y_was_categorical, test_size=0.2)

这是在我的情况下工作的代码