Wsrtjtyk

I've created a function that tests multiple IF statements given the data in the 'Name' column.

Criteria 1: If 'Name' is blank, return the 'Secondary_Name'. However, if 'Secondary_Name' is also blank, return the 'Third_Name'.

Criteria 2: If 'Name' == 'GENERAL', return the 'Secondary_Name'. However, if 'Secondary_Name' is also blank, return the 'Third_Name'

Else: Return the 'Name'

def account_name(row):

    if row['Name'] == None and row['Secondary_Name'] == None:

        return row['Third_Name']



    elif row['Name'] == 'GENERAL':

        if row['Secondary_Name'] == None:

            return row['Third_Name']



    else:

        return row['Name']

I've tried == None, == np.NaN, == Null, .isnull(), == '', == '0'.

Nothing seems to replace the empty values to what I want.

Edit:

Example of DF

Depending on column dtype, NULLs dont interact well with NaN for whatever reason.
None would mean that the field is blank, which it is not since you have "a value" for a given dtype.

The simple way to ensure you can identify NaNs is to see if the field is equal to itself

def isNaN(value):

if value != value: 

    return True

else:

    return False

And to provide an example:

df = pd.DataFrame(data={'ClientId':[1,2,3,4] , 'SomeNULLs':['main','main',None,None], 'NewNULLs':[1,None,0,1]})



df['Test'] = df.NewNULLs.apply(isNaN)

The resulting dataset should be

   ClientId SomeNULLs  NewNULLs   Test

0         1      main       1.0  False

1         2      main       NaN   True

2         3      None       0.0  False

3         4      None       1.0  False

Consider this df

df = pd.DataFrame({'Name':['a', 'GENERAL', None],'Secondary_Name':['e','f',None], 'Third_Name':['x', 'y', 'z']})



    Name    Secondary_Name  Third_Name

0   a       e               x

1   GENERAL f               y

2   None    None            z

Since you are writing the function in python, you can use is None

def account_name(row):



    if (row['Name'] is None or row['Name'] == 'GENERAL') and (row['Secondary_Name'] is None):

        return row['Third_Name']



    elif row['Name'] is None or row['Name'] == 'GENERAL':

        return row['Secondary_Name']



    else:

        return row['Name']



df['Name'] = df.apply(account_name, axis = 1)

You get

    Name    Secondary_Name  Third_Name

0   a       e               x

1   f       f               y

2   z       None            z

You can get same output using pandas and nested np.where

cond1 = (df['Name'].isnull()) | (df['Name'] == 'GENERAL')

cond2 = (cond1) & (df['Secondary_Name'].isnull())

np.where(cond2, df['Third_Name'], np.where(cond1, df['Secondary_Name'], df['Name']))