Pandas dataframe add 2 dataframes

Question

I need to be able to add the values of two dataframes with the same structure together and form a new dataframe as a result.

e.g. DF1 + DF2 = DF3

DF1
+------------+----+----+----+
|    date    |  A |  B |  C |
+------------+----+----+----+
| 2017-01-01 | 24 | 15 |  4 |
| 2017-01-02 | 31 | 10 | 12 |
| 2017-01-03 |  9 | 47 |  3 |
+------------+----+----+----+

DF2
+------------+----+----+----+
|    date    |  A |  B |  C |
+------------+----+----+----+
| 2017-01-01 |  4 | 12 | 63 |
| 2017-01-02 | 23 |  0 | 31 |
| 2017-01-03 | 61 | 22 | 90 |
+------------+----+----+----+

DF3
+------------+----+----+----+
|    date    |  A |  B |  C |
+------------+----+----+----+
| 2017-01-01 | 28 | 27 | 67 |
| 2017-01-02 | 64 | 10 | 43 |
| 2017-01-03 | 70 | 69 | 93 |
+------------+----+----+----+

I've been trying to work out how to do this but i'm getting a TypeError

TypeError: unsupported operand type(s) for +: 'datetime.date' and 'datetime.date'

when trying to do:

df3 = df1.add(df2, fill_value=0)

I'm sure i'm missing something simple as it appears to be trying to add the first columns (which is a date and the column I want to match on to add together the values for all other columns) but any help would be greatly appreciated.

if you take out the date column and add it later on, it should work. is this a possiblity given the structure of your data? it will work in the above case. — clog14
– clog14, Commented Apr 11, 2017 at 16:21
The dates need to remain linked to the data otherwise the data isn't any good. I suppose basically what I need is to add up the corresponding columns based on the date matching between the datafrmes — Richard Newman
– Richard Newman, Commented Apr 11, 2017 at 16:26
Another option would be using concat, groupby and sum: df3 = pd.concat([df1, df2]).groupby('date').sum().reset_index() — Jan Trienes
– Jan Trienes, Commented Apr 11, 2017 at 16:35

ASGM · Accepted Answer · 2017-04-11 16:31:04Z

3

You want the date columns to be indices, not normal columns:

df3 = df1.set_index('date').add(df2.set_index('date'), fill_value=0).reset_index()

You use set_index() so that the date columns becomes indices. If you don't want the final dataframe to be date-indexed, you can use reset_index() at the end as @MaxU suggests.

edited Apr 11, 2017 at 16:31

answered Apr 11, 2017 at 16:26

ASGM

11.5k1 gold badge37 silver badges54 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

MaxU - stand with Ukraine Over a year ago

i'd also add .reset_index() at the end ;-)

Richard Newman Over a year ago

@ASGM, thanks very much for your help. Adding the indices was the solution I needed :)

Collectives™ on Stack Overflow

Pandas dataframe add 2 dataframes

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related