How to append a series value using if...else in Python-CodePudding

I have a data frame with around 10k Observation and few features. Manually in excel file I can create a new column based on given data i.e.

 = IF(A = 0, 1, B)

where A, B are basically Column name.

Output of above formula would be appended on new column i.e. C.

I have used below Python code :

C = []

for x in df['A']:
    if x == 0:
        C.append(1)
    else:
        C.append(df['B'])

But the above output is giving me "unhashable type: 'Series'" Error.

Anyway to correct the above code

CodePudding user response：

Don't use a loop, use np.where:

import numpy as np
# as new column
df['C'] = np.where(df['A'].eq(0), 1, df['B'])
# as list
C = np.where(df['A'].eq(0), 1, df['B']).tolist()

or mask:

# as new column
df['C'] = df['B'].mask(df['A'].eq(0), 1)

# as list
C = df['B'].mask(df['A'].eq(0), 1).to_list()

CodePudding user response：

You are trying to append the whole column - you need to append the value. Try:

c = []
for a, b in zip(df['A'], df['B']):
    c.append(1 if a == 0 else b)