Replace a value in a column if that value appears only once

Question

I have a dataframe, I want to replace the values in one column by "other" if the value count of that value in that column is exactly 1

i        Food_group
0          Flake
1          Flake
2          Flake
3         Almond
4          Drink
5          Drink
6          Flake

I have tried,

data["food_group"] = data.food_group.apply(lambda x: "other" if x.value_counts()==1 else x)

I got error

AttributeError: 'str' object has no attribute 'value_counts'

Simon Larsson · Accepted Answer · 2019-04-21T09:45:21.167

3

It is easier if you store the value count separately to avoid redoing it inside the apply loop. You can do it like this:

food_count = data["food_group"].value_counts()
data["food_group"] = data["food_group"].apply(lambda x: "other" if food_count[x]==1 else x)

edited Apr 21 '19 at 09:45

answered Apr 21 '19 at 09:40

Simon Larsson

Have you tried the code? it does not work – KHAN irfan Apr 21 '19 at 09:43
Test it now. :) Tried on another dataset and forgot to change back to "food_group" – Simon Larsson Apr 21 '19 at 09:45
Tested it on your data now, so it should work. – Simon Larsson Apr 21 '19 at 09:51

1 Answers1