设置新列时如何不获取 SettingWithCopyWarning

Question

I have an annoying problem with SettingWithCopyWarning and I don't seem to get rid of the warning. I have a df users like this

user id     cloud_files 
    1              0
    2              8
    3              1

I would like to add a boolean column of cloud users

user id     cloud_files     cloud_user  
    1              0           false
    2              8           true  
    3              1           true  

I have tried to do it but I keep getting the SettingWithCopyWarning.

users['cloud_user'] = np.where(users['cloud_files']== 0, True, False)

users['cloud_user'] = users['cloud_files']!= 0

users.loc[:,'cloud_user'] = users['cloud_files']!= 0

I have tried copy and reset index but I just can't get rid of the warning. Is there some way to do it without getting this warning or do I just need to set the warning off?

Answer 1

You can use properties of booleans to directly create your new column because bool(num) will evaluate to True as long the num is non-zero

df['cloud_user'] = df['cloud_files'].astype(bool)

Result:

>>> df = pd.DataFrame({'user_id':[1,2,3],'cloud_files':[0,8,1]})
>>> df['cloud_user'] = df['cloud_files'].astype(bool)
>>> df
   user_id  cloud_files  cloud_user
0        1            0       False
1        2            8        True
2        3            1        True