-1
Phone_no  Date     Problem
 1     2020-01-19   G
 1     2020-01-19   A
 1     2020-01-27   B
 1     2020-01-28   C
 1     2020-01-28   H
 2     2020-01-19   T
 3     2020-01-04   U
 3     2020-01-22   P
 4     2020-01-12   E
 5     2020-01-01   G
 5     2020-01-11   A
 2     2020-01-31   I
 2     2020-01-31   E

**I want to apply groupby on Phone_no,Date and want data should bein dataframe as mention below **

Phone_no  Date     Problem

1     2020-01-19    G,A
1     2020-01-27     B
1     2020-01-28    C,H
2     2020-01-19     T
2     2020-01-31    I,E
3     2020-01-04     U
3     2020-01-22     P
4     2020-01-12     E
5     2020-01-01     G
5     2020-01-11     A
Sajid Choudhary
  • 11
  • 2

1 Answers1

1

You can groupby "Phone_no" and "Date" and use agg to apply a lambda function that joins "Problem" values in each group:

out = df.groupby(['Phone_no','Date'])['Problem'].agg(lambda x: ','.join(x)).reset_index()

or in a simpler way (thanks @Corralien):

out = df.groupby(['Phone_no','Date'], as_index=False)['Problem'].apply(','.join)

if you want tuples:

out = df.groupby(['Phone_no','Date'])['Problem'].apply(tuple).reset_index()

Output:

   Phone_no        Date Problem
0         1  2020-01-19     G,A
1         1  2020-01-27       B
2         1  2020-01-28     C,H
3         2  2020-01-19       T
4         2  2020-01-31     I,E
5         3  2020-01-04       U
6         3  2020-01-22       P
7         4  2020-01-12       E
8         5  2020-01-01       G
9         5  2020-01-11       A