merge and sum two dataframes where columns match python pandas

Question

I might be approaching this is the wrong way. If I have two dataframes a and b:

Dataframe A:

a b c
1 2 4
1 6 5
1 8 7

and Dataframe B:

a b c d
1 2 4 9
1 6 5 7
1 8 7 10

And I want to join them, but cruically sum the columns where the column name matchs, and keep the column names that don't match. So the final joined table would look like:

a   b   c   d
2   4   8   9
2   12  10  7
2   16  14  10

Note: The indexs would always match (ie the same number of records in the same order)

Answer 1

You could call add on the larger df passing in the other df and call fillna :

In [18]:
df1.add(df).fillna(df1)

Out[18]:
   a   b   c   d
0  2   4   8   9
1  2  12  10   7
2  2  16  14  10

Another way is to use combine_first :

In [20]:
df1.add(df).combine_first(df1)

Out[20]:
   a   b   c   d
0  2   4   8   9
1  2  12  10   7
2  2  16  14  10

Answer 2

You can align the dataframe, fill NAN with zero and perform simple addition.

d = pd.DataFrame(data, columns=['a','b','c'])
d2 = pd.DataFrame(data2, columns=['a','b','c','d'])
d, d2 = d.align(d2, fill_value=0)
sum_df = d + d2

In [23]: d
Out[23]: 
   a  b  c  d
0  1  2  4  0
1  1  6  5  0
2  1  8  7  0

In [24]: d2
Out[24]: 
   a  b  c   d
0  1  2  4   9
1  1  6  5   7
2  1  8  7  10

In [25]:sum_df
Out[25]: 
   a   b   c   d
0  2   4   8   9
1  2  12  10   7
2  2  16  14  10