df_a and df_b are two dataframes that looks like following
df_a
df_b
df_a A B C D E x1 Apple 0.3 0.9 0.6 x1 Orange
You can still achieve this with a left join which is very ideal. See below:
final_df = pd.merge(df_a, df_b[['A', 'B_new','F']], how="left", left_on=['A', 'B'], right_on=['A', 'B_new']);