Scikit learn 当我尝试调用score方法时，spark_sklearn包的GridSearchCV失败

我试图使用

spark\u sklearn

包中的

GridSearchCV

而不是

sklearn

来利用

spark

但当我调用估计器的

score

方法时，它失败了

我从中获取了示例代码

代码如下所示：

def example_ppl():
    import numpy as np
    from sklearn import linear_model, decomposition, datasets
    from sklearn.pipeline import Pipeline
    # from sklearn.model_selection import GridSearchCV
    from spark_sklearn import GridSearchCV
    logistic = linear_model.LogisticRegression()

    pca = decomposition.PCA()
    pipe = Pipeline(steps=[('pca', pca), ('logistic', logistic)])

    digits = datasets.load_digits()
    X_digits = digits.data
    y_digits = digits.target

    n_components = [20, 40, 64]
    Cs = np.logspace(-4, 4, 3)
    # Create spark context
    spark_session =  SparkSession.builder.appName('test').getOrCreate()
    sc = spark_session.sparkContext

    estimator = GridSearchCV(sc,
                         estimator=pipe,
                         param_grid=dict(pca__n_components=n_components,
                              logistic__C=Cs))
    print(type(estimator))
    estimator.fit(X_digits, y_digits)
    # print(estimator.cv_results_)
    estimator.score(X_digits,y_digits) 

它抛出一个错误，如下所示：

File "D:/Python_Project/test/sklearn_pyspark.py", line 72, in example_ppl
estimator.score(X_digits,y_digits)
File "D:\PyEnvs\test\lib\site-packages\sklearn\model_selection\_search.py", line 436, in score
score = self.scorer_[self.refit] if self.multimetric_ else self.scorer_
AttributeError: 'GridSearchCV' object has no attribute 'multimetric_'

是因为

spark\u sklearn

的问题还是我的代码中遗漏了什么