Python - 将字符串 'NA' 视为 Null 或 NaN 来执行 fillna()

Question

I have a dataframe:

mydf = pd.DataFrame({'foo': ['bar', 'NA', 'Bar'], 'bar': ['NA', 'foo', 'NA']})

I would like to fill the 'NA' values using the fillna() function. However, fillna() does not detect 'NA' as Null/NaN values.

Would I have to convert all 'NA' to NaN? Or is there a way I can treat 'NA' as NaN in my dataframe?

Answer 1

You can use replace function

mydf.replace('NA', <value_you_want>, inplace=True)

Or, first, replace 'NA' with NaN and the use fillna method

import numpy as np
mydf.replace('NA', np.nan, inplace=True)
mydf.fillna(<value_you_want>)

Answer 2

您需要将所有 'NA' 转换为 NumPy.nan 值，因为 python 检测到 'None' 而不是 'NA'。