Need to fill the missing value-using mode method(most frequently occuring value).

Output needs as below:

If the product is A, the value should be 3 and for B, it is 4.

CodePudding user response：

You can use a custom groupby.apply. Note that there can be several modes, we take the first available one here:

df['Another'] = (df.groupby('Product')['Another']
                   .apply(lambda g: g.fillna(g.mode()[0], downcast='infer'))
                )

output:

  Product  Another
0       A        3
1       A        3
2       A        3
3       A        3
4       B        4
5       B        4
6       B        4
7       B        4

Alternative

If you expect a single valid value per group, use groupby.transform('first') instead:

df['Another'] = df.groupby('Product')['Another'].transform('first')

CodePudding user response：

Solution

  import pandas as pd

  dict = {"Product":["A", "A", "A", "A", "B", "B", "B", "B"],   "Another" : 
  [3,3,None,None,4,4,None,None]}

  df  = pd.DataFrame(dict)
  null_data = df[df.isnull().any(axis=1)]

  mode = df['Another'].mode()

  df.loc[df.Product == "A", "Another"] = mode[0]
  df.loc[df.Product == "B", "Another"] = mode[1]