如何将数据框转换为字典?

Question

我有以下数据框架:

Group SubGroup
0  GroupA       A1
1  GroupA       A2
2  GroupA       A3
3  GroupB       B1
4  GroupB       B2
5  GroupC       C1

如何在字典中转换这个数据框?我期望的输出如下:

{'GroupA': ['A1', 'A2', 'A3'], 'GroupB': ['B1', 'B2'], 'GroupC': ['C1']}

Answer 1

您可以使用groupby，然后to_dict功能:

In [2]: df = pd.DataFrame({'group':['A','A','A','B','B','C'], 'subgrroup':['A1','A2','A3','B1','B2','C1']})
In [3]: df
Out[3]: 
group subgrroup
0     A        A1
1     A        A2
2     A        A3
3     B        B1
4     B        B2
5     C        C1
In [5]: df2 = df.groupby('group')['subgrroup'].apply(list)
In [6]: df2
Out[6]: 
group
A    [A1, A2, A3]
B        [B1, B2]
C            [C1]
Name: subgrroup, dtype: object
In [7]: df2.to_dict()
Out[7]: {'A': ['A1', 'A2', 'A3'], 'B': ['B1', 'B2'], 'C': ['C1']}

Answer 2

df.groupby('Group')['SubGroup'].apply(list).to_dict()

Answer 3

您可以尝试使用字典理解pd.DataFrameGroupBy.keys:

>>> {k: df.loc[v, 'SubGroup'].tolist() for k, v in df.groupby('Group').groups.items()}
{'GroupA': ['A1', 'A2', 'A3'], 'GroupB': ['B1', 'B2'], 'GroupC': ['C1']}
>>> 

或者试试pd.SeriesGroupBy.agg:

>>> dict(df.groupby('Group')['SubGroup'].agg(list))
{'GroupA': ['A1', 'A2', 'A3'], 'GroupB': ['B1', 'B2'], 'GroupC': ['C1']}
>>> 

使用to_dict:

更好

>>> df.groupby('Group')['SubGroup'].agg(list).to_dict()
{'GroupA': ['A1', 'A2', 'A3'], 'GroupB': ['B1', 'B2'], 'GroupC': ['C1']}
>>>