Pandas Dataframe groupby Display

pandas, python

Solution

You can nearly do this with a plain ol' `sum`:

In [11]: df.groupby('B').sum()
Out[11]:
              A
B
200  text1text2
300  text1text2
400       text1
500       text2
600       text1

You could use an aggregate with a `join`:

In [12]: df.groupby('B').agg(lambda x: ', '.join(x.values))
Out[12]:
                A
B
200  text1, text2
300  text1, text2
400         text1
500         text2
600         text1

Problem

Lets say I have a DataFrame like following. ``` A B 0 text1 200 1 text2 200 2 text1 400 3 text2 500 4 text1 300 5 text1 600 6 text2 300 ``` I want to print following output ``` A B 0 text1,text2 200 2 text1 400 3 text2 500 4 text1,text2 300 5 text1 600 ``` There is no order,I just want to take text labels of column "A" for matching values in column "B". This has to be done with `df.groupby` as I know. Any way no success with my efforts yet. Hope you get my question.

Original source