Pivoting column while retaining all other columns-CodePudding

I have many columns in a table, but only one column that needs to be pivoted with its values. It looks like this:

OrderNumber  Item  YearMonth  Total
1            1     2019_01    20
1            2     2019_01    40
1            1     2019_02    30
2            1     2019_02    50

The resulting output should be:

OrderNumber  Item  2019_01    2019_02   Total
1            1     60         30        20
1            2     60         30        40
1            1     60         30        30
2            1     0          50        50

Basically, sum all the total for each month's order number while retaining all columns. Is there a way to do this? I'm using Pandas

CodePudding user response：

IIUC, you need a pivot_table merge:

out = (df
  .merge(df.pivot_table(index='OrderNumber', columns='YearMonth',
                        values='Total', aggfunc='sum', fill_value=0),
         on='OrderNumber')
  #.drop(columns='YearMonth') # uncomment to drop unused 'YearMonth'
  )

Output:

   OrderNumber  Item YearMonth  Total  2019_01  2019_02
0            1     1   2019_01     20       60       30
1            1     2   2019_01     40       60       30
2            1     1   2019_02     30       60       30
3            2     1   2019_02     50        0       50