So my table currently looks like this:

Based on this I want to add a new column

I am really confused as to how this would be. I am able to write the first instance of a roll no as "Other subject to xyz" but cannot seem to understand how the following values will carry over. Thank you

CodePudding user response：

Using df.groupby() and .transform() methods with lambda

df["Details"] = (df.groupby("Roll No")
                    .Subject.transform(lambda x: x.shift()
                    .add(" to ").fillna("Other subject to ")   x))
print(df)

   Roll No    Subject                   Details
0        1       Math     Other subject to Math
1        1    Science           Math to Science
2        2    History  Other subject to History
3        2  Geography      History to Geography
4        2   Politics     Geography to Politics
5        3    Finance  Other subject to Finance

CodePudding user response：

You can use only pandas.DataFrame.shift on columns "Roll No", "Subject" then use numpy.where.

import numpy as np

ss = df['Subject'].shift()
sr = df['Roll No'].shift()
r1 = ss   ' to '   df['Subject']
r2 = 'Other subject to '   df['Subject']
df['res'] = np.where(sr==df['Roll No'], r1, r2)
print(df)

Output:

   Roll No    Subject                       res
0        1       Math     Other subject to Math
1        1    Science           Math to Science
2        2    History  Other subject to History
3        2  Geography      History to Geography
4        2   Politics     Geography to Politics
5        3    FInance  Other subject to FInance