在 Kaggle 內核上使用 train_test_Split 方法時顯示錯誤。 幫助表示贊賞

Question

任何人都可以解釋什么是錯誤以及如何避免？ 相同的代碼實際上適用於我的 Jupyter 和 Spider 環境。 我也不明白為什么這不適用於 kaggle 內核。

下面是我的代碼。

# 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)

錯誤詳情：

        ---------------------------------------------------------------------------
    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

可能是因為您的目標變量的類型不是分類的，您必須將其更改為標簽。

Answer 2

這是一個尚未解決的錯誤。

為了使代碼工作，您的響應變量不應是分類的。

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)

這是在我的情況下工作的代碼