分组多个属性并将结果/保存传递给_excel

Question

df;

OpCode         PYear OperationStartDate
0    MCI6AF2   1993           2017-03-24
1    MCI6BF2   1994           2017-03-24
2    APJ5LFC   1975           2017-03-13
3    MQL0AF3   1986           2017-03-16
4    BMR3KFC   1982           2017-03-15
5    BRM1NFC   1981           2017-03-15
6    LNP5GF2   1987           2017-03-16
7       EAS5   1999           2017-03-22
8    HRK1DSC   1971           2017-03-17
9    HRK2FSC   1972           2017-03-17
10   NMI6BF2   1994           2017-03-22
11   NMI6AF2   1994           2017-03-20
12   HHY1JF1   1996           2017-02-01

我能够使用一个属性“ opcode”进行分组，以在“ OperationStartDate”中获得最小日期；将最短日期传递给联接，然后使用将输出保存到excel；

df2 = df.reset_index()
df3 = df2.loc[df2.groupby('OpCode')['LastOperationDate'].idxmin()] #Try df3 = df2.loc[df2.groupby(['OpCode','PYEAR'])['LastOperationDate'].idxmin()]

重命名df3“ LastOperationDate”列

df3.rename(columns={'LastOperationDate':'EarliestStartDate'}, inplace=True)


df3.drop(df3.columns[[0,4,1,5,6]], axis=1, inplace=True)


df6 = pd.merge(left = df, right = df3, how ='left', on ='OpCode')

writer = ExcelWriter('C:\Current Projects\Mapping\Python\Process Outputs\Example1321.xlsx')
df6.to_excel(writer,'Sheet1',index=False)
writer.save()

问题是，如果我对多个属性OpCode和PYear进行分组，则会得到既无法传递给联接也无法保存为excel的结果。在这种情况下，我正在使用以下替代代码；

df3 =df['Operation Start Date'].groupby([df['OpCode'], df['PYear']]).min()

df4= df.groupby(['OpCode', 'PYear'])['Operation Start Date'].apply(lambda x : x.min())

我的结果；

 Op Code                P Year                
ANN3EF1                1994                     2017-08-29
APJ5LFC                1975                     2017-03-13
APJ6EFC                1977                     2017-10-09
APJ6GFC                1976                     2017-09-04
APJ6HFC                1975                     2017-09-07
APJ6JFC                1977                     2017-07-27
AYN4RFC                1968                     2017-11-13
AYN6CF1                1984                     2017-03-31
AYN6CF3                1984                     2017-08-29
AYN6DF3                1985                     2017-09-05
AZZ7AF1                2004                     2017-08-04

有人知道我如何将其传递给以下联接；

df6 = pd.merge(left = df, right = df3, how ='left', on ='OpCode')

或在输出的日期中添加列标题

或将其保存到excel

Answer 1

您不需要加入，只需在这里进行var ReportFolderListUtil = { confirmBox2: { hide: function () { } }, ：

transform