Python 数组与 json

Question

假设我有下一个数据框。

import pandas as pd 

# initialize list of lists 
data = [[10,'A',2], [11,'A',4], [11,'B',7], [11,'C',4], [12,'A',9], [12,'C',8], [12,'D',5]] 

# Create the pandas DataFrame 
df = pd.DataFrame(data, columns = ['User', 'Pref','Score']) 

# print dataframe. 
df 

User    Pref    Score
0   10  A   2
1   11  A   4
2   11  B   7
3   11  C   4
4   12  A   9
5   12  C   8
6   12  D   5

df.groupby('User')['Pref'].apply(lambda g: g.to_json(orient='records')).reset_index()

User    Pref
0   10  ["A"]
1   11  ["A","B","C"]
2   12  ["A","C","D"]

如何获得一个新的数据帧，其中偏好和分数在每个用户的级别上分开。

User    Pref
0   10  [{"A",2}]
1   11  [{"A",4},{"B",7},{"C",4}]
2   12  [{"A",9},{"C",8},{"D",5}]

Answer 1

你可以这样做

 df_ = df.groupby('User').apply(lambda x : [{a,b} for (a,b) in zip(x.Pref,
                                                x.Score)]).to_frame('Pref')

输出

df_
Out[11]: 
                          Pref
User                          
10                    [{2, A}]
11    [{4, A}, {B, 7}, {4, C}]
12    [{9, A}, {8, C}, {D, 5}]