CODEWITHSUNDEEP
pythonpandasdataframe
N. Hinkel
N. Hinkel

Reputation: 15

How to sum a column of one pandas dataframe to each column of another dataframe?

Situation:

Both dataframes (df1 and df2) have the same three indices, e.g. "A", "B", "C". df1 and df2 differ in the number of their columns. All cells in df1 and df2 are filled with data of the type float.

df1:

   AA    BB   CC   DD
A  28.0   2.3  2.0  113
B  24.0   2.4  2.4  116
C  33.5  10.0  4.0    0

df2:

    AAA   BBB   CCC    DDD    EEE
A  1.01  4.04  7.07  10.10  13.13
B  2.02  5.05  8.08  11.11  14.14
C  3.03  6.06  9.09  12.12  15.15

Goal:

After the column chosen from df2 (e.g. "BBB") is summed with each column of df1, the results should be stored in a new dataframe (df_new). df_new should be of the format of df1 (number of columns and rows) and have the same column-names and indices as df1.

new_df:

    AA      BB      CC      DD
A   32.04   6.34    6.04    117.04
B   29.05   7.45    7.45    121.05
C   39.56   16.06   10.06   6.06

My approach:

To sum the 3rd column of df2 with each column of df1, I tried:

df_new = df1.add(df2.iloc[:,2])

Result:

As desired, df_new matches df1 in structure and labels. The problem is: All cells show "NaN" instead of the desired sums as floats.

df_new:

    AA  BB  CC  DD
A   NaN NaN NaN NaN
B   NaN NaN NaN NaN
C   NaN NaN NaN NaN

Upvotes: 0

Views: 2483

Answers (1)

Zero
Zero

Reputation: 76917

Here's one way, using add

In [293]: df1.add(df2['BBB'], axis=0)
Out[293]:
      AA     BB     CC      DD
A  32.04   6.34   6.04  117.04
B  29.05   7.45   7.45  121.05
C  39.56  16.06  10.06    6.06

Upvotes: 1

Related Questions