[FIXED] pandas problem when assigning value using loc

Issue

So what is happening is the values in column B are becoming NaN. How would I fix this so that it does not override other values?

import pandas as pd
import numpy as np
# %%
# df=pd.read_csv('testing/example.csv')
data = {
    'Name' : ['Abby', 'Bob', 'Chris'],
    'Active' : ['Y', 'Y', 'N'],
    'A' : [89, 92, np.nan],
    'B' : ['eye', 'hand', np.nan],
    'C' : ['right', 'left', 'right'] 
}

df = pd.DataFrame(data)
df.loc[((df['Active'] =='N') & (df['A'].isna())), ['A', 'B']] = [99, df['C']]
df 

What I want the results to be is:

Name	Active	A	B	C
Abby	Y	89.0	eye	right
Bob	Y	92.0	hand	left
Chris	N	99	right	right

Solution

You can implement this by creating a boolean mask using the condition: Where the 'Active' column is 'N' and the 'A' column has missing values (np.nan). We can then use this mask to filter rows in the DataFrame.

First, we will replace the values in column 'A' with 99 for rows where the mask condition is True. Then, replace the values in column 'B' with the corresponding values from column 'C' where the mask condition is True.

Following is the modified code:

import pandas as pd
import numpy as np

data = {
    'Name' : ['Abby', 'Bob', 'Chris'],
    'Active' : ['Y', 'Y', 'N'],
    'A' : [89, 92, np.nan],
    'B' : ['eye', 'hand', np.nan],
    'C' : ['right', 'left', 'right'] 
}

df = pd.DataFrame(data)
mask = (df['Active'] =='N') & (df['A'].isna())

df.loc[mask, 'A'] = 99
df.loc[mask, 'B'] = df.loc[mask, 'C']

print(df)

Now, the DataFrame will be updated correctly. Following is the output:

    Name Active     A      B      C
0   Abby      Y  89.0    eye  right
1    Bob      Y  92.0   hand   left
2  Chris      N  99.0  right  right

Answered By - Bilesh Ganguly