比较数据框值错误中的 2 列

Question

I have a data-frame df1 which looks like:

  ID  myColA  myColB myColC
0  A       1       5     13
1  B      -2       6     14
2  C       3      -7     15
3  D       4       8     16

I am trying to add a new column myColD that is filled on the result of the following:

myColD = ((myColA > 0 and myColB <0) or (myColA < 0 and myColB > 0)),0,myColA)

where row value in myColA is above 0 and value in myColB is below 0 or vice versa return 0, else return the myColA value.

So my desired output would be:

  ID  myColA  myColB myColC myColD
0  A       1       5     13      1
1  B      -2       6     14      0  
2  C       3      -7     15      0
3  D       4       8     16      4

Here is my code:

df1 = pd.DataFrame({'ID': ['A', 'B', 'C', 'D'],
    'myColA': [1, -2, 3, 4],
    'myColB': [5, 6, -7, 8],
    'myColC': [9, 10, 11, 12]},
     index=[0, 1, 2, 3])

df1['myColD'] = np.where(((df1.myColA > 0) & (df1.myColB < 0)) or ((df1.myColA < 0) & (df1.myColB > 0)), df1.myColA, 0)

I am however getting a Value Error:

ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

Answer 1

You need bitwise operators for this. So use a bitwise or : | instead of or .

Also note that you could simplify this by checking where the product between both columns results in a negative value, and set the corresponding values to 0 with mask :

df1['myColD'] = df1.myColA.mask(df1.myColA.mul(df1.myColB).lt(0), 0)

---

print(df1)

   ID  myColA  myColB  myColC  myColD
0  A       1       5       9       1
1  B      -2       6      10       0
2  C       3      -7      11       0
3  D       4       8      12       4
​

Answer 2

point to the columns like this

import pandas as pd
import numpy as np
df1 = pd.DataFrame({'ID': ['A', 'B', 'C', 'D'],
    'myColA': [1, -2, 3, 4],
    'myColB': [5, 6, -7, 8],
    'myColC': [9, 10, 11, 12]},
     index=[0, 1, 2, 3])

df1['myColD'] = np.where(((df1['myColA'] > 0) & (df1['myColB'] < 0)) | ((df1['myColA'] < 0) & (df1['myColB'] > 0)), df1['myColA'], 0)
print(df1)