I'm applying this to a more extensive data set, but the idea is the same. Let me elaborate.
I have the following data frame:
list = [('a' , 'X', 'good '),
('a', 'Y', 'ok '),
('b', 'Y', 'nice '),
('b', 'Z', 'bad '),
('c', 'X', 'ugly ')]
df = pd.DataFrame(list, columns = ['user', 'item', 'text'])
I group all the user text into one column, and I do the same for the items. (I've completed this step)
This creates two new data frames called df_user and df_item
df_user = **user | text**
'a' | good ok
'b' | nice bad
'c' | ugly
df_item = **item | text**
'X' | good ugly
'Y' | ok nice
'Z' | bad
And the final step would be adding the text from df_user and df_item to df.
df should look like this:
df = ** user | item | user_text | item_text**
a | X | good ok | good ugly
a | Y | good ok | ok nice
b | Y | nice bad | ok nice
b | Z | nice bad | bad
c | X | ugly | good ugly
My question is, how to add the grouped text data from df_user and df_item into df, as shown above?