python - 属性错误: module "sklearn.utils" has no attribute "_joblib" when inheriting class `sklearn.ensemble.BaggingClassifier.`

Question

我需要提取在 sklearn.ensemble.BaggingClassifier 中训练的每个模型的概率。这样做的原因是为了估计 XGBoostClassifier 模型的不确定性。

为此，我创建了一个继承自 sklearn.ensemble.BaggingClassifier 的扩展类，并添加了一个允许获取这些概率的新方法。请注意，此问题与 ModuleNotFoundError: No module named 'sklearn.utils._joblib' 不同

我在下面展示了到目前为止我已经实现的代码片段:

必要的模块

from sklearn.ensemble import BaggingClassifier
from sklearn.ensemble.base import _partition_estimators
from sklearn.utils import check_array
from sklearn.utils.validation import check_is_fitted
import sklearn.utils as su

继承自BaggingClassifier的子类

class EBaggingClassifier(BaggingClassifier):
    """
    Extends the class BaggingClassifier fromsklearn

    """

    def __init__(self,
                 base_estimator=None,
                 n_estimators=10,
                 max_samples=1.0,
                 max_features=1.0,
                 bootstrap=True,
                 bootstrap_features=False,
                 oob_score=False,
                 warm_start=False,
                 n_jobs=1,
                 random_state=None,
                 verbose=0):

        super().__init__(
                 base_estimator,
                 n_estimators,
                 max_samples,
                 max_features,
                 bootstrap,
                 bootstrap_features,
                 oob_score,
                 warm_start,
                 n_jobs,
                 random_state,
                 verbose)

允许计算每个估计量概率的新方法定义如下。

    def predict_proball(self, X):
        """
        Computes the probability of each individual estimator

        Parameters
        ----------
        X : {array-like, sparse matrix} of shape = [n_samples, n_features]
            The training input samples. Sparse matrices are accepted only if
            they are supported by the base estimator.

        Returns
        -------
        p : array of shape = [n_samples, n_classes]
            The class probabilities of the input samples. The order of the
            classes corresponds to that in the attribute `classes_`.
        """

        check_is_fitted(self, "classes_")
        # Check data
        X = check_array(
            X, accept_sparse=['csr', 'csc'], dtype=None,
            force_all_finite=False
        )

        if self.n_features_ != X.shape[1]:
            raise ValueError("Number of features of the model must "
                             "match the input. Model n_features is {0} and "
                             "input n_features is {1}."
                             "".format(self.n_features_, X.shape[1]))

        # Parallel loop
        n_jobs, n_estimators, starts = _partition_estimators(self.n_estimators,
                                                             self.n_jobs)

        all_proba = su._joblib.Parallel(n_jobs=n_jobs, verbose=self.verbose,
                             **self._parallel_args())(
            su._joblib.delayed(BaggingClassifier._parallel_predict_proba)(
                self.estimators_[starts[i]:starts[i + 1]],
                self.estimators_features_[starts[i]:starts[i + 1]],
                X,
                self.n_classes_)
            for i in range(n_jobs))

        return all_proba

我使用 XGBoostClassifier 作为基本估计器实例化此类:

base_estimator = XGBoostClassifier(**params)
estimator = EBaggingClassifier(base_estimator=base_estimator, max_samples=0.8, n_estimators=10)

然后使用 estimator.fit(X, y) 进行估计器，其中 X 和 y 是pandas.DataFrame 对象。当我尝试运行 estimator.predict_proball(X) 我得到

>>> estimator.predict_proball(X)
AttributeError: module 'sklearn.utils' has no attribute '_joblib'

有人知道为什么会发生这种情况吗？查看BaggingClassifier script函数“sklearn.utils._joblib”应该可用。

仅供引用:

>>> sklearn.__version__
'0.19.2'

Answer 1

问题出在您的 scikit-learn 版本上。版本'0.19.2'没有_joblib，可以引用here 。或者您可以使用以下命令进行检查:

dir(su)

您需要更新scikit-learn，最新版本有_joblib，可以引用here .

您在版本'0.20.2'中获得以下内容:

>>> dir(su)
['Bunch', 'DataConversionWarning', 'IS_PYPY', 'Memory', 'Parallel', 'Sequence', 
'_IS_32BIT', '__all__', '__builtins__', '__cached__', '__doc__', '__file__', 
'__loader__', '__name__', '__package__', '__path__', '__spec__', '_joblib', 
'_show_versions', 'as_float_array', 'assert_all_finite', 'axis0_safe_slice', 
'check_X_y', 'check_array', 'check_consistent_length', 'check_random_state', 
'check_symmetric', 'class_weight', 'column_or_1d', 'compute_class_weight', 
'compute_sample_weight', 'cpu_count', 'delayed', 'deprecate', 'deprecated', 
'deprecation', 'effective_n_jobs', 'fixes', 'gen_batches', 'gen_even_slices', 
'get_chunk_n_rows', 'get_config', 'hash', 'indexable', 'indices_to_mask', 
'is_scalar_nan', 'issparse', 'msg', 'murmurhash', 'murmurhash3_32', 'np', 
'numbers', 'parallel_backend', 'platform', 'register_parallel_backend', 
'resample', 'safe_indexing', 'safe_mask', 'safe_sqr', 'shuffle', 'struct', 
'tosequence', 'validation', 'warnings']

您可以按如下方式更新scikit-learn:

pip install -U scikit-learn