无法在列表理解中索引 dict答案

【问题标题】：Can't index a dict in a list comprehension无法在列表理解中索引 dict
【发布时间】：2016-07-20 17:24:35
【问题描述】：

我正在尝试从dicts 的list 中提取list。列表中的每个dict 都有一个键值映射。我想为 list 包含的每个 dict 中映射的单个值创建一个 list。

查看下面的代码和输出，以交互形式编写。

>>> print(str(param_to_outputs_dict_dict))
{('max_features', 1558): {'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=1558, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}, 
('max_features', 7713): {'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=7713, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
>>> print()

在上面我们看到param_to_outputs_dict_dict 是两个条目的dict。每个条目的键是tuple，值是dict。

接下来，我使用列表解析来提取值的list（dict 类型）

>>> map_list = [output_dict for output_dict in list(param_to_outputs_dict_dict.values())]
>>> print(str(map_list))
[{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=1558, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}, 
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=7713, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}]
>>> print()

我们看到这也有效。我们有一个 list 包含两个 dicts，它们是 param_to_outputs_dict_dict 中包含的值。

现在，我不提取字典，而是尝试提取每个字典中由值 'classifier' 映射的值。

>>> classifier_list = [output_dict['classifier'] for output_dict in list(param_to_outputs_dict_dict.values())]
>>> print(str(classifier_list))


---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-176-b6f236339c69> in <module>()
      8 print(str(map_list))
      9 print()
---> 10 classifier_list = [output_dict['classifier'] for output_dict in list(param_to_outputs_dict_dict.values())]
     11 print(str(classifier_list))
     12 print()

<ipython-input-176-b6f236339c69> in <listcomp>(.0)
      8 print(str(map_list))
      9 print()
---> 10 classifier_list = [output_dict['classifier'] for output_dict in list(param_to_outputs_dict_dict.values())]
     11 print(str(classifier_list))
     12 print()
TypeError: 'DecisionTreeClassifier' object is not subscriptable

但是，这不起作用。错误消息让我认为 Python 将 output_dict 解释为 DecisionTreeClassifier 类型，但我不知道为什么。我已经玩了几个小时这个问题，但完全被难住了。

我怀疑解决方案对其他人来说可能非常明显！提前致谢！

编辑：原始帖子中缺少关键信息

为了使问题更简短、更清晰，我将打印结果截断为仅包含任何集合的前两个元素。然而，在意识到问题之后（见下面我的回答），当给出完整列表时，问题的原因就非常明显了。见下文，特别是列表的最后一个元素。

>>> for value in list(param_to_outputs_dict_dict.values()):
>>>     print(str(value))
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=10,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=100,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-100, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=3, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-10000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=7791, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=77, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=100000,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=1558, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=1,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=3895, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=7713, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=7011, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight='balanced', criterion='gini',
        max_depth=None, max_features=None, max_leaf_nodes=None,
        min_samples_leaf=1, min_samples_split=2,
        min_weight_fraction_leaf=0.0, presort=False, random_state=None,
        splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='random')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-1000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=779, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=10000,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-1, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=1000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=10, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=1000,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=10000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=6232, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=100, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=88, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-10, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=100000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
{'classifier': DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=-100000, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')}
DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
        max_features=None, max_leaf_nodes=None, min_samples_leaf=1,
        min_samples_split=2, min_weight_fraction_leaf=0.0,
        presort=False, random_state=None, splitter='best')

【问题讨论】：

我个人看不出有什么问题。我不知道你的 DecisionTreeClassifier 类是什么样的，但是用我编的这个假的，它似乎运行良好：ideone.com/uzNw17
@DavidWhite 感谢您为我运行它。嗯，这很令人困惑！仅供参考，我在我的代码中使用来自 scikit-learn 的 sklearn.tree.DecisionTreeClassifier。

标签： dictionary scikit-learn list-comprehension python-3.5

【解决方案1】：

查看问题中打印的完整列表，我们可以看到最后一个元素不是包含单个 str 到 DecisionTreeClassifier 映射的 dict 类型，而是简单的 DecisionTreeClassifier 类型。

因此，以下代码：

[output_dict['classifier'] for output_dict in list(param_to_outputs_dict_dict.values())]

将list(param_to_outputs_dict_dict.values()) 的每个元素分配给output_dict 在最后一个元素处分配DecisionTreeClassifier。

因此，在该迭代中，语句 output_dict['classifier'] 尝试下标 DecisionTreeClassifier 对象，不是 dict。这会产生问题中给出的TypeError。

【讨论】：

【解决方案2】：

您正在做一些多余的事情。以下是实现的方法。

[v["classifier"] for v in param_to_outputs_dict_dict.values()]

list(dict.values()) 和[v for v in dict.values()] 是同一个东西，那么[v for v in list(dict.values())] 的语义是什么。在同一行中，print(str(x)) 到底是什么？ print(x) 输出 x.__str__()。当我说冗余时，这就是我的意思。

【讨论】：

在 Python 3 中，dict.values() 返回一个视图，而不是一个列表。该视图不可索引，并且如果基础 dict 更改，则会更改。因此，您的实现不太线程安全。
另外，我更喜欢显式转换数据；说str(x)（x 的字符串表示形式）对我来说看起来更清楚，而不是将非字符串类型传递给print 并依靠它来读懂我的想法。