Pandas：使用通配符匹配从给定值返回列名？

Question

import pandas as pd

data = pd.read_csv('transactions(1).csv')
userTransactionsDF = pd.DataFrame(data, columns=['Date', 'Original Description', 'Amount', 'Category',
                                                             'Account Name'])
print(userTransactionsDF.head(1))

# output
#         Date Original Description  ...             Category             Account Name
# 0  4/14/2023            PETCO 304  ...  Pet Food & Supplies  DISCOVER IT CHROME CARD

data2 = pd.read_csv('recommended_categories.csv')
recommendedCategoriesDF = pd.DataFrame(data2)

print(recommendedCategoriesDF .head(1))

# output
#    Unnamed: 0  HOUSING TRANSPORT   FOOD  ... INCOME ENTERTAINMENT SERVICES Other
# 0           0      NaN     shell  qdoba  ...    NaN      nintendo      NaN   NaN

如果

userTransactionsDF['Original Description'][0]

与

recommendedCategoriesDF

中的任何值匹配 在

recommendedCategoriesDF

中的值名称的两侧使用 '.+' 通配符，然后我想返回与通配符匹配的值所在的列名称

例如如果

['Original Description'][0]

输出

'POS WD  Nintendo CA877180016   8'

然后想要它匹配

'nintendo'

在recommendedCategoriesDF并返回列名

'ENTERTAINMENT'

我尝试了以下将推荐的类别数据框转换为字典并返回键，如果该值与字典中的任何值匹配并带有通配符。我想我离基地很远……

        for j in userTransactionsDF['Original Description']:
            for p in userTransactionsDF['Amount']:
                print('original description = ' + j)
                #need to implement re to implement wildcard in below logical statement
                if (i for i in categories_dict if re.search('.+' + str(categories_dict[i]) + '.+', j)):
                    value = {i for i in categories_dict if re.search('.+' + str(categories_dict[i]) + '.+', j)}
                    print('value = ' + str(value))
                    categorized_Transactions.loc[len(categorized_Transactions)] = pandas.Series({value: p})
                else: categorized_Transactions.loc[len(categorized_Transactions)] = pandas.Series({'Other': p})