处理NaN数据的问题

Question

我正在尝试根据同名数据创建新文件。现在我只在列表上工作，这些列表存储匹配的行名。 我的问题是，所有的NaN值不是存储在一个列表中。我猜NaN总是不等于NaN，因为没有价值

import pandas as pd

df = pd.read_excel('/Users/jedknauer/code/Python xlsx/OPOS-Liste Kreditoren zum 31.08.2021.xlsx')
(df.sort_values("Unnamed: 2",ascending = False)["Unnamed: 2"])

#df1 = df.loc[df["Unnamed: 2"]== "Kluxen & Co. Immobilien"][]


#for index, row in df.iterrows():
#    print(index, row["Unnamed: 2"])

new_objects = []

for index, row in df.iterrows():
    b = row["Unnamed: 2"]
    is_new = False
    
    if (index == 0):
        new_objects.append([b,[]])
    for i  in new_objects:
        
        if (b == i[0]):
            is_new = False
            i[-1].append(b)
            break; 
        if(b != i[0] ):
            is_new = True
    if (is_new ==  True):
        new_objects.append([b,[]])
        is_new == False

print(new_objects)

这是新_对象的输出：

[['Beschriftung', ['Beschriftung']], ['Korg & More', ['Korg & More', 'Korg & More']],
 ['B.& K. Baumgärtel GmbH', ['B.& K. Baumgärtel GmbH', 'B.& K. Baumgärtel GmbH', 'B.& K. Baumgärtel GmbH']],
 ['hallo PC', ['hallo PC']], ['Bundesanzeiger Verlag', []], ['Steinway & Sons', ['Steinway & Sons']],
 [nan, []], [nan, []], ['Mach Ernst', ['Mach Ernst', 'Mach Ernst']], ['Aaron Padelt', []], [nan, []], ['EP Service', []],
 [nan, []], ['Kluxen & Co. Immobilien', ['Kluxen & Co. Immobilien', 'Kluxen & Co. Immobilien', 'Kluxen & Co. Immobilien',
                                         'Kluxen & Co. Immobilien', 'Kluxen & Co. Immobilien', 'Kluxen & Co. Immobilien',
                                         'Kluxen & Co. Immobilien']],
 [nan, []], [nan, []], [nan, []], ['Universaldrucker', []], ['EMD Musik s.a.', ['EMD Musik s.a.']],
 [nan, []], ['KSC', ['KSC', 'KSC', 'KSC', 'KSC']], ['Dumrath & Fassnacht', ['Dumrath & Fassnacht']],
 [nan, []], ['Telekom Deutschland GmbH', ['Telekom Deutschland GmbH', 'Telekom Deutschland GmbH']],
 [nan, []], ['KRAVAG', []], [nan, []], [nan, []], [nan, []], [nan, []], [nan, []], [nan, []], [nan, []],
 [nan, []], ['DomainFactory GmbH', []], ['Dickau, Claudia', []], ['Klavier Atelier Blankenese', []],
 [nan, []], ['Vattenfall Europe Sales GmbH', ['Vattenfall Europe Sales GmbH']], ['Diverse Kreditoren', []], [nan, []]]