如果日期介于 pandas 中的 2 个日期之间，则查找每个 ID 的值之和

Question

data={'id':[1,1,1,1,2,2,2,2],
      'date1':[datetime.date(2016,1,1),datetime.date(2016,1,2),datetime.date(2016,1,2),datetime.date(2016,1,4),
               datetime.date(2016,1,2),datetime.date(2016,1,4),datetime.date(2016,1,3),datetime.date(2016,1,1)],
      'date2':[datetime.date(2016,1,5),datetime.date(2016,1,3),datetime.date(2016,1,5),datetime.date(2016,1,5),
               datetime.date(2016,1,4),datetime.date(2016,1,5),datetime.date(2016,1,4),datetime.date(2016,1,1)],
      'score1':[5,7,3,2,9,3,8,3],
      'score2':[1,3,0,5,2,20,7,7]}
df=pd.DataFrame.from_dict(data)

数据框 df 看起来像这样：

   id       date1       date2  score1  score2
0   1  2016-01-01  2016-01-05       5       1
1   1  2016-01-02  2016-01-03       7       3
2   1  2016-01-02  2016-01-05       3       0
3   1  2016-01-04  2016-01-05       2       5
4   2  2016-01-02  2016-01-04       9       2
5   2  2016-01-04  2016-01-05       3      20
6   2  2016-01-03  2016-01-04       8       7
7   2  2016-01-01  2016-01-01       3       7

另一个包含ID、usetdate的数据框UF

   id      usetdate  
0   1  2016-01-01    
1   1  2016-01-03    
2   2  2016-01-04    
3   2  2016-01-02    

如果我通过的日期（usetdate）在date1和date2之间，我想针对UF的每个ID找到score1和score2的总和

   id      usetdate  score1  score2
0   1  2016-01-01      5       1
1   1  2016-01-03     17       9
2   2  2016-01-04     20       29
3   2  2016-01-02     9        2

Answer 1

您可以首先通过

Series

 创建包含所有日期时间的 

date_range，将索引与

Series

和

DataFrame.join

中的值交换为原始的最后聚合

sum

:

s = pd.concat([pd.Series(r.Index,pd.date_range(r.date1, r.date2)) for r in df.itertuples()])
s = pd.Series(s.index, index=s, name='usetdate')

df = df.drop(['date1','date2'],axis=1).join(s).groupby(['id','usetdate'], as_index=False).sum()
print (df)
   id   usetdate  score1  score2
0   1 2016-01-01       5       1
1   1 2016-01-02      15       4
2   1 2016-01-03      15       4
3   1 2016-01-04      10       6
4   1 2016-01-05      10       6
5   2 2016-01-01       3       7
6   2 2016-01-02       9       2
7   2 2016-01-03      17       9
8   2 2016-01-04      20      29
9   2 2016-01-05       3      20

编辑：

L = [(i, d, s1, s2) for i, d1, d2, s1, s2 in df.values for d in pd.date_range(d1, d2)]
df = (pd.DataFrame(L, columns=['id','usetdate','score1','score2'])
        .groupby(['id','usetdate'], as_index=False).sum())
print (df)
   id   usetdate  score1  score2
0   1 2016-01-01       5       1
1   1 2016-01-02      15       4
2   1 2016-01-03      15       4
3   1 2016-01-04      10       6
4   1 2016-01-05      10       6
5   2 2016-01-01       3       7
6   2 2016-01-02       9       2
7   2 2016-01-03      17       9
8   2 2016-01-04      20      29
9   2 2016-01-05       3      20

编辑：

您可以在聚合之前使用左连接

merge

值：

df1['userdate'] = pd.to_datetime(df1['userdate'])
print (df1)
   id   userdate
0   1 2016-01-01
1   1 2016-01-03
2   2 2016-01-04
3   2 2016-01-02

L = [(i, d, s1, s2) for i, d1, d2, s1, s2 in df.values for d in pd.date_range(d1, d2)]
df = (pd.DataFrame(L, columns=['id','userdate','score1','score2'])
        .merge(df1)
        .groupby(['id','userdate'], as_index=False)
        .sum())
print (df)
   id   userdate  score1  score2
0   1 2016-01-01       5       1
1   1 2016-01-03      15       4
2   2 2016-01-02       9       2
3   2 2016-01-04      20      29

编辑1：

您可以过滤列表理解中转换为元组的值：

df1['userdate'] = pd.to_datetime(df1['userdate'])
print (df1)
   id   userdate
0   1 2016-01-01
1   1 2016-01-03
2   2 2016-01-04
3   2 2016-01-02

a = [tuple(x) for x in df1.values]
print (a)
[(1, Timestamp('2016-01-01 00:00:00')), (1, Timestamp('2016-01-03 00:00:00')), 
 (2, Timestamp('2016-01-04 00:00:00')), (2, Timestamp('2016-01-02 00:00:00'))]
L = [(i, d, s1, s2) for i, d1, d2, s1, s2 in df.values 
                    for d in pd.date_range(d1, d2) 
                    if (i, d) in a]
df = (pd.DataFrame(L, columns=['id','userdate','score1','score2'])
        .groupby(['id','userdate'], as_index=False)
        .sum())
print (df)
   id   userdate  score1  score2
0   1 2016-01-01       5       1
1   1 2016-01-03      15       4
2   2 2016-01-02       9       2
3   2 2016-01-04      20      29

Answer 2

import datetime
import pandas as pd

data={'id':[1,1,1,1,2,2,2,2],
      'date1':[datetime.date(2016,1,1),datetime.date(2016,1,2),datetime.date(2016,1,2),datetime.date(2016,1,4),
               datetime.date(2016,1,2),datetime.date(2016,1,4),datetime.date(2016,1,3),datetime.date(2016,1,1)],
      'date2':[datetime.date(2016,1,5),datetime.date(2016,1,3),datetime.date(2016,1,5),datetime.date(2016,1,5),
               datetime.date(2016,1,4),datetime.date(2016,1,5),datetime.date(2016,1,4),datetime.date(2016,1,1)],
      'score1':[5,7,3,2,9,3,8,3],
      'score2':[1,3,0,5,2,20,7,7]}
df=pd.DataFrame.from_dict(data)

data={'id':[1,1,2,2],
      'date':[datetime.date(2016,1,1),datetime.date(2016,1,2),datetime.date(2016,1,2),datetime.date(2016,1,4)]}

df1=pd.DataFrame.from_dict(data)

data1=[]
x=[]
def agg(df,df1):
    for i in range(1,2):
        x=list(df1.id)
        y=list(df1.date)
        data1= df[df.id==x[i]]
        data2=data1[data1.date1 > y[i]] 
        data3=data2[data1.date2 < y[i]]
        data4=data3.groupby(['id']).agg({"score1":sum})
        x.append(data4)
    return data4

agg(df,df1)

请尝试这个

Answer 3

df1.sql.set_alias('df1').join(df2.sql.set_alias('df2'),condition='df1.id=df2.id').filter("usetdate between date1 and date2").aggregate("df2.id,usetdate,sum(score1) score1,sum(score2) score2")

┌───────┬────────────┬────────┬────────┐
│  id   │  usetdate  │ score1 │ score2 │
│ int64 │  varchar   │ int128 │ int128 │
├───────┼────────────┼────────┼────────┤
│     1 │ 2016-01-03 │     15 │      4 │
│     1 │ 2016-01-01 │      5 │      1 │
│     2 │ 2016-01-04 │     20 │     29 │
│     2 │ 2016-01-02 │      9 │      2 │
└───────┴────────────┴────────┴────────┘