Wrangling Pandas Dataframe

I have a large dataframe that looks like below.

PERIOD    Value
1         30
1         40
1         50
2         60
2         75
2         80
3         90
3         100
3         200

I want to have the DataFrame in following format without using for loop.

PERIOD     Value
1          30
2          60
3          90
1          40
2          75
3         100
1          50
2          80
3         200

The order of entries has to be maintained. Any help is appreciated.

>Solution :

You can use groupby.cumcount to (stably) sort the values then use the index to reorder:

order = df.groupby('PERIOD').cumcount().sort_values(kind='stable').index

out = df.loc[order]

Variant using the key parameter of sort_values:

out = df.sort_values(by='PERIOD', kind='stable',
                     key=lambda x: x.groupby(x).cumcount())

Output:

   PERIOD  Value
0       1     30
3       2     60
6       3     90
1       1     40
4       2     75
7       3    100
2       1     50
5       2     80
8       3    200