Group a groupby

Question

Hi all so I have the following column in my dataframe:

     LC_REF
1   DT 16 2C
2   DT 16 2C
3   DT 16 2C

1    DT 16 3C
6    DT 16 3C
3    DT 16 3C
7    DT 16 3C

0    DT 17 1C
0    DT 17 1C
0    DT 17 1C
1    DT 17 1C

0    DT 17 2C
2    DT 17 2C
4    DT 17 2C
5    DT 17 2C

4   DT 17 3C
1   DT 17 3C
8   DT 17 3C

And I'm wondering if its possible to organize every group of LC_REF's such that the index is in ascending order:

     LC_REF
1   DT 16 2C
2   DT 16 2C
3   DT 16 2C

1    DT 16 3C
3    DT 16 3C
6    DT 16 3C
7    DT 16 3C

0    DT 17 1C
0    DT 17 1C
0    DT 17 1C
1    DT 17 1C

0    DT 17 2C
2    DT 17 2C
4    DT 17 2C
5    DT 17 2C

1   DT 17 3C
4   DT 17 3C
8   DT 17 3C

so far I have tried:

 df.groupby('LC_REF').apply(pd.DataFrame.sort_index,'LC_REF',1)
 grouped = df.groupby('LC_REF').sum().reset_index()

but that does not give the correct output.

Answer 1

Or As my comment , that is the 1st good way to go . But if you you do want to using groupby try this

df.groupby('LC_REF',as_index=False).apply(pd.DataFrame.sort_index)

Answer 2

df = pd.read_table(StringIO("""
id   LC_REF
1   DT 16 2C
2   DT 16 2C
3   DT 16 2C
1    DT 16 3C
6    DT 16 3C
3    DT 16 3C
7    DT 16 3C
0    DT 17 1C
0    DT 17 1C
0    DT 17 1C
1    DT 17 1C
0    DT 17 2C
2    DT 17 2C
4    DT 17 2C
5    DT 17 2C
4   DT 17 3C
1   DT 17 3C
8   DT 17 3C"""), sep = '   ')

df.sort_values(['LC_REF', 'id'],ascending=True)

Answer 3

I think better is use reset_index for column from index, then sort_values and last set_index:

df = df.reset_index().sort_values(['LC_REF','index']).set_index('index').rename_axis(None)
print (df)
     LC_REF
1  DT 16 2C
2  DT 16 2C
3  DT 16 2C
1  DT 16 3C
3  DT 16 3C
6  DT 16 3C
7  DT 16 3C
0  DT 17 1C
0  DT 17 1C
0  DT 17 1C
1  DT 17 1C
0  DT 17 2C
2  DT 17 2C
4  DT 17 2C
5  DT 17 2C
1  DT 17 3C
4  DT 17 3C
8  DT 17 3C