0

So I have a data frame called df. It looks like this.

0 1 2
1 2 3
4 5 6
7 8 9

I want to sum up the columns and divide the sum of the columns by the sum of the rows. So for example: row 1, column 0: (1+4+7)/(1+2+3) row 2, column 0: (1+4+7)/(4+5+6)

so on and so forth.

so that my final result is like this.

0 1 2
2 2.5 3
0.8 1 1.2
0.5 0.625 0.75

How do I do it in python using pandas and dataframe?

indian cluster
  • 1

2 Answers2

2

You can also do it this way:

import numpy as np
a = df.to_numpy()
b = np.divide.outer(a.sum(0),a.sum(1)) 
# divide is a ufunc(universal function) in numpy.
# All ufunc's support outer functionality
out = pd.DataFrame(b, index=df.index, columns=df.columns)

output:

     0      1     2
0  2.0  2.500  3.00
1  0.8  1.000  1.20
2  0.5  0.625  0.75
MSS
  • 3,306
  • 1
  • 19
  • 50
1

You can use the underlying numpy array:

a = df.to_numpy()
out = pd.DataFrame(a.sum(0)/a.sum(1)[:,None],
                   index=df.index, columns=df.columns)

output:

     0      1     2
0  2.0  2.500  3.00
1  0.8  1.000  1.20
2  0.5  0.625  0.75
mozway
  • 194,879
  • 13
  • 39
  • 75