1

我正在尝试编写一个可以合并两个字典（文本文件！）的程序。这些词典由名词和动词组成，这些名词和动词已被另一个程序从不同的语料库中索引（然后放入文本文件中）。这是这些字典的形式：

dict1 = {'strawberry': [['eat', 1]], 'family-member': [['look up', 1]], 'mall': [['search', 1]]}
dict2 = {'strawberry': [['eat', 1]], 'family-member': [['lose', 1]], 'ovation': [['receive', 1]], 'mall': [['build', 1]]}

如您所见，它们是带有键的字典，在值列表中包含列表。现在我试图得到这样的输出：

finaldict = {'strawberry': [['eat', 2]], 'family-member': [['look up', 1]['lose',1]], 'mall': [['search', 1]['build', 1]], 'ovation': [['receive', 1]]

到目前为止，我已经能够像这样（在字符串中）合并 dict1 和 dict2：

{'strawberry': [['eat', 1]], 'family-member': [['look up', 1]], 'mall': [['search',
1]], 'strawberry': [['eat', 1]], 'family-member': [['lose', 1]], 'ovation':
[['receive', 1]], 'mall': [['build', 1]]}

我将此字符串转换为带有下一条语句的字典： finaldict = eval(str1) 它将整个内容转换为字典，当我询问 finaldict 的类型时它也这么说，但它不会看到像 [ ['eat', 1]] 作为值或任何东西。我需要这个，所以我可以遍历每个项目并计算它与哪个动词一起出现的次数。

4

2 回答 2

1 
from collections import Counter

dict1 = {'strawberry': [['eat', 1]], 'family-member': [['look up', 1]], 'mall': [['search', 1]]}
dict2 = {'strawberry': [['eat', 1]], 'family-member': [['lose', 1]], 'ovation': [['receive', 1]], 'mall': [['build', 1]]}
result = {k: Counter(dict(v)) for k, v in dict1.items()}
for k, v in dict2.items():
    result.setdefault(k, Counter()).update(dict(v))

result = {k: [list(x) for x in v.items()] for k, v in result.items()}
于 2012-06-14T22:18:45.583 回答
0

没有什么太花哨的，只是打破它。

from collections import defaultdict

dict1 = {'strawberry': [['eat', 1]], 'family-member': [['look up', 1]], 'mall': [['search', 1]]}
dict2 = {'strawberry': [['eat', 1]], 'family-member': [['lose', 1]], 'ovation': [['receive', 1]], 'mall': [['build', 1]]}
keys = set(dict2.keys()).union(dict1.keys())

final = {}
for k in keys:
    d1val = dict1.get(k, [])
    d2val = dict2.get(k, [])

    resd = defaultdict(lambda: 0)

    for word, count in d1val:
        resd[word] += count

    for word, count in d2val:
        resd[word] += count

    final[k] = [list(i) for i in resd.items()]
于 2012-06-14T22:24:32.070 回答