Pandas: Count Unique combinations from multiple columns and write amount in new column

Asked Oct 29 '18 at 13:57

Active Oct 29 '18 at 13:57

Viewed 21 times

Lets say i have following dataframe df:

col1    col2 
-------------
apple   apple
melon   apple
apple   apple
pear    melon

Now, how do i count all combinations and write it into a new column? Like this.

col1    col2   amount
---------------------
apple   apple    2
melon   apple    1
pear    melon    1

asked Oct 29 '18 at 13:57

kxell2001

Use `df.groupby(['col1', 'col2']).size().reset_index(name='amount')` – jezrael Oct 29 '18 at 13:58
Does it also work with NaN entries or empty string? Because the count doesnt sum up with the total row number. – kxell2001 Oct 29 '18 at 14:15
NaNs values are excluded, emoty string working nice. – jezrael Oct 29 '18 at 14:15

0 Answers0