通过最小索引号对Pandas数据框组进行排序，然后基于第三列对组中的所有其他列进行重新排序

Question

I have a dataframe like this:

import numpy as np
import pandas as pd

columns=['Order', 'Group_code', 'Grade', 'Contextual_info']

data = np.array([np.arange(6)]*4).T
mydf = pd.DataFrame(data, columns=columns)

mydf.Order = [1,2,3,4,5,6]
mydf.Group_code = ['group99','group2','group2','group2','group12','group12']
mydf.Grade = [80,0,60,80,85,70]
mydf.Contextual_info = [5,4,3,2,1,0]

mydf

  Order Group_code Grade Contextual_info 0 1 group99 80 5 1 2 group2 0 4 2 3 group2 60 3 3 4 group2 80 2 4 5 group12 85 1 5 6 group12 70 0 

Which is ordered by Order . I want to preserve the ordering of Group_code by Order , thus the column values in Group_code should not change.

However, within each Group_code group, I want to order the rows descending by Grade . Finally, I will replace Order with a new vector of integers 1... n, such that it is still 1, 2, 3, 4, 5, 6 in this example.

Desired result:

  Order Group_code Grade Contextual_info 1 group99 80 5 2 group2 80 2 3 group2 60 3 4 group2 0 4 5 group12 85 1 6 group12 70 0 

Answer 1

Use

In [677]: mydf.Grade = (mydf.groupby('Group_code')['Grade']
                            .transform(pd.Series.sort_values, ascending=False))

In [678]: mydf
Out[678]:

  Order Group_code Grade 0 1 group99 80 1 2 group2 80 2 3 group2 60 3 4 group2 0 4 5 group12 85 5 6 group12 70