How to multiply each row in pandas dataframe by a different value

Question

I am trying to multiply each row of a pandas dataframe by a different value and wondering what the best way to do this is.

For example if I have the following dataframe:

import numpy as np
import pandas as pd
df = pd.DataFrame(np.random.randn(2, 3))
df
     0          1           2
0   -1.283316   0.849488    1.936060
1   -2.078575   -0.871570   -0.970261

I want to multiply each element of each row by a different in a list or array: vals = [1, 100].

In this example I want each item in the first row to be multiplied by 1 and each item in the second row to be multiplied by 100

the result should therefore be:

     0          1           2
0   -1.283316   0.849488    1.936060
1   -207.8575   -87.1570    -97.0261

I have tried:

df * vals
df.multiply(vals)
df.multiply(vals, axis=1)

None of which work, although I was not expecting them too, based on my understanding of what that code should do. I can’t figure out a concise way to do this with pandas.

Asked By: johnchase

||

Source

Answer 1

The method is mul:

df.mul([1, 100], axis=0)
Out[17]: 
            0          1         2
0   -1.198766  -1.340028  1.990843
1  113.890468 -68.177755 -9.060228

Answered By: ayhan

Answer 2

The most pandastic way is mul on axis=0. That said, to make * work, we need a column array where the number of rows match the number of rows of the frame. So the following work.

vals = np.array([1, 100])
df1 = df * vals[:, None]

# or to modify the original frame
df *= vals[:, None]

Answered By: cottontail

How to multiply each row in pandas dataframe by a different value

Question:

Answers: