np.where not implemented on new column-CodePudding

Adding column3 with:

df['column3'] = np.where((df['column2']) == 1, '1', 0)

to the 2 columns existing table don't feed the 1's and the result is column3 full with 0's

column1   column2   column3 
0         1         0
1         0         0
0         0         0
0         1         0
1         0         0

CodePudding user response：

FTR, the problem seems to be that the column has string values, meaning no value will be equal to 1. Use "1" instead:

df['column3'] = np.where(df['column2'] == "1", "1", 0)

CodePudding user response：

So the following code is working for me. It is essentially the same as your code.

import pandas as pd
import numpy as np


d = {'column1': [0, 1, 0, 0, 1], 'column2': [1, 0, 0, 1, 1], 'column3': [0, 0, 0, 0, 0]}
df = pd.DataFrame(data=d)

df['column3'] = np.where((df['column2']) == 1  , '1', 0)
```

The output is:
[enter image description here][1]


  [1]: https://i.stack.imgur.com/sgodK.png

So please check the datatype of "column2", if it is a string the condition may be false.