sum up two pandas dataframes with different indexes element by element

5
votes

I have two pandas dataframes, say df1 and df2, of some size each but with different indexes and I would like to sum up the two dataframes element by element. I provide you an easy example to better understand the problem:

dic1 = {'a': [3, 1, 5, 2], 'b': [3, 1, 6, 3], 'c': [6, 7, 3, 0]}
dic2 = {'c': [7, 3, 5, 9], 'd': [9, 0, 2, 5], 'e': [4, 8, 3, 7]}
df1 = pd.DataFrame(dic1)
df2 = pd.DataFrame(dic2, index = [4, 5, 6, 7])

so df1 will be

   a  b  c
0  3  3  6
1  1  1  7
2  5  6  3
3  2  3  0

and df2 will be

   c  d  e
4  7  9  4
5  3  0  8
6  5  2  3
7  9  5  7

now if type

df1 + df2

what I get is

    a   b   c   d   e
 0 NaN NaN NaN NaN NaN
 1 NaN NaN NaN NaN NaN
 2 NaN NaN NaN NaN NaN
 3 NaN NaN NaN NaN NaN
 4 NaN NaN NaN NaN NaN
 5 NaN NaN NaN NaN NaN
 6 NaN NaN NaN NaN NaN
 7 NaN NaN NaN NaN NaN

How can I make pandas understand that I want to sum up the two dataframe just element by element?

1
pythonnumpypandasdataframe
If you don't care about the indexes, df1.values + df2.values - chrisb

1 Answers

10
votes

UPDATE: much better solution from piRSquared:

In [39]: df1 + df2.values
Out[39]:
    a   b   c
0  10  12  10
1   4   1  15
2  10   8   6
3  11   8   7

Old answer:

In [37]: df1.values + df2.values
Out[37]:
array([[10, 12, 10],
       [ 4,  1, 15],
       [10,  8,  6],
       [11,  8,  7]], dtype=int64)

In [38]: pd.DataFrame(df1.values + df2.values, columns=df1.columns)
Out[38]:
    a   b   c
0  10  12  10
1   4   1  15
2  10   8   6
3  11   8   7