Sklearn 管道抛出 ValueError: too many values to unpack

Question

我正在尝试创建一个 sklearn 管道，该管道将首先提取文本中的平均字长，然后使用StandardScaler.

定制变压器

class AverageWordLengthExtractor(BaseEstimator, TransformerMixin):

    def __init__(self):

        pass

    def average_word_length(self, text):

        return np.mean([len(word) for word in text.split( )])

    def fit(self, x, y=None):

        return self

    def transform(self, x , y=None):

        return pd.DataFrame(pd.Series(x).apply(self.average_word_length))

我的目标是实现这一目标。X 是一个带有文本值的熊猫系列。这行得通。

    extractor=AverageWordLengthExtractor()

    print(extractor.transform(X[:10]))

    sc=StandardScaler()

    print(sc.fit_transform(extractor.transform(X[:10])))

我为此创建的管道是。

pipeline = Pipeline([('text_length', AverageWordLengthExtractor(), 'scale', StandardScaler())])

但pipeline.fit_transform()产生以下错误。

Traceback (most recent call last):

  File "custom_transformer.py", line 48, in <module>

    main()

  File "custom_transformer.py", line 43, in main

    'scale', StandardScaler())])

  File "/opt/conda/lib/python3.6/site-packages/sklearn/pipeline.py", line 114, in __init__

    self._validate_steps()

  File "/opt/conda/lib/python3.6/site-packages/sklearn/pipeline.py", line 146, in _validate_steps

    names, estimators = zip(*self.steps)

ValueError: too many values to unpack (expected 2)

Answer 1

您的括号在错误的位置/创建管道时缺少括号，应该是元组列表：pipeline = Pipeline([&nbsp; &nbsp;('text_length', AverageWordLengthExtractor()),&nbsp;&nbsp; &nbsp;('scale', StandardScaler())])

Answer 2

我认为您需要将fit_transform方法添加到您的课程中AverageWordLengthExtractor。