python - 将系列转换为字典？

Question

我有一个应用了 groupby 函数的数据集

data.groupby(by=['version','sentiment']).score.count()

输出是:

version  sentiment       
10.1.2   negative        3.0
         positive        8.0
10.2.0   negative        1.0
10.4.2   negative        1.0
10.4.3   negative        1.0
10.4.6   positive        1.0
10.5.1   negative        1.0
10.5.2   positive        1.0
10.5.5   negative        2.0
11.0     negative        1.0
11.2     negative        3.0
         positive        1.0
11.3.1   negative        1.0

不，我想将其转换为字典形式，输出应类似于:

{'10.1.2' : {
      'positive' : 3,
      'negative' : 8
},
'10.2.0' : {
}

Answer 1

我认为你只需要:

s.unstack().to_dict('index')
#.unstack(level = 0).to_dict()

s 是您的系列

data.groupby(by=['version','sentiment']).score.count().unstack().to_dict()

<小时/>

我使用 DataFrame 来解决您的问题

print(df)
        i1        i2  col
0   10.1.2  negative  3.0
1   10.1.2  positive  8.0
2   10.2.0  negative  1.0
3   10.4.2  negative  1.0
4   10.4.3  negative  1.0
5   10.4.6  positive  1.0
6   10.5.1  negative  1.0
7   10.5.2  positive  1.0
8   10.5.5  negative  2.0
9     11.0  negative  1.0
10    11.2  negative  3.0
11    11.2  positive  1.0
12  11.3.1  negative  1.0


df.set_index(['i1','i2'])['col'].unstack().to_dict('index')
#{'10.1.2': {'negative': 3.0, 'positive': 8.0},
# '10.2.0': {'negative': 1.0, 'positive': nan},
# '10.4.2': {'negative': 1.0, 'positive': nan},
# '10.4.3': {'negative': 1.0, 'positive': nan},
# '10.4.6': {'negative': nan, 'positive': 1.0},
# '10.5.1': {'negative': 1.0, 'positive': nan},
# '10.5.2': {'negative': nan, 'positive': 1.0},
# '10.5.5': {'negative': 2.0, 'positive': nan},
# '11.0': {'negative': 1.0, 'positive': nan},
# '11.2': {'negative': 3.0, 'positive': 1.0},
# '11.3.1': {'negative': 1.0, 'positive': nan}}