TypeError：__ init __（）得到了一个意外的关键字参数＆＃39; shuffle＆＃39;

Question

在python 2.6.6中运行我的代码时出现此错误。在Python 3.4.3中运行时没有问题

usr/lib64/python2.6/site-packages/sklearn/feature_selection/univariate_selection.py:319: UserWarning: Duplicate scores. Result may depend on feature ordering.There are probably duplicate features, or you used a classification score for a regression task.
      warn("Duplicate scores. Result may depend on feature ordering."
    Traceback (most recent call last):
      File "classification.py", line 31, in <module>
        main()
      File "classification.py", line 15, in main
        tm.optimaltrain(conf)
      File "/axp/gabm/npscnnct/dev/getThemes/textminer/textminer/classify.py", line 121, in optimaltrain
        score = self.cv(X,y,model)
      File "/axp/gabm/npscnnct/dev/getThemes/textminer/textminer/classify.py", line 140, in cv
        skf = cross_validation.StratifiedKFold(y, n_folds=self.cv_folds, shuffle=True)
    TypeError: __init__() got an unexpected keyword argument 'shuffle'

代码：

  def cv(self, X, y, model):
    y_true = []
    y_pred = []
    skf = cross_validation.StratifiedKFold(y, n_folds=self.cv_folds, shuffle=True)
    for train_index, test_index in skf:
      X_train, X_test = X[train_index], X[test_index]
      y_train, y_test = y[train_index], y[test_index]
      model.fit(X_train, y_train)
      y_pred += list(model.predict(X_test))
      y_true += list(y_test)

但是当我从代码中删除Shuffle=True时，其运行正常。我使用的模块是scipy 0.11.0，nltk 2.0.1，sklearn 0.14.1

请指教。 感谢

Answer 1

以下是for(j<=1; j<=i ; j=2*j);// performs no action 版本0.14）的来源：https://github.com/scikit-learn/scikit-learn/blob/0.14.X/sklearn/cross_validation.py#L391

我已链接到sklearn上 init 的实际行 - 这表明没有StratifiedKFold个关键字参数。

升级到具有shuffle的v 0.15（如此处所示：https://github.com/scikit-learn/scikit-learn/blob/0.15.X/sklearn/cross_validation.py#L399）。

我将假设shuffle上的sklearn版本为Py3？

Answer 2

在sklearn 0.14中，cross_validation.StratifiedKFold()没有关键字参数shuffle。显然，它只是在以后的版本中添加（实际上是0.15）。

您可以在分层之前更新sklearn或自行更改输入（例如，使用random.shuffle()）。