pandas关键字提取_从pandas数据帧列中提取关键字单词，但不提取嵌套关键字

我有一个Pandas DataFrame，其中包含两个自由格式的文本列，人们在其中描述其车辆的型号或装饰级别(例如：LE、1LT、RS、SS等)。在这些列中，有些人将只拥有模型(例如：LE)，其他人将添加附加文本(例如：2dr Convertible SS w/2SS)。此外，模型级别具有一定的层次结构，即SS&lt；1SS&lt；2SS。在

我想提取这些模型或修剪级别，并在我的数据帧中创建一个新列(例如：1ls=1ls，ZL-1=ZL1，等等)# the model can be stored in either 'submodel' or 'trim'

data = [{'SubModel': 'SS-EDITION', 'Trim': 'SS-EDITION(MANUAL 6 SPEED) Coupe 2-Door'},

{'SubModel': 'ZL1', 'Trim': 'ZL1 Coupe 2-Door'},

{'SubModel': 'N/A', 'Trim': 'SS TRANSFORMER'},

{'SubModel': '1LT RS AUTO BLUETOOTH REAR CAM', 'Trim': 'N/A'},

{'SubModel': 'N/A', 'Trim': 'LS'},

{'SubModel': 'Camaro SS', 'Trim': 'Camaro SS'},

{'SubModel': 'Dusk Edition', 'Trim': 'N/A'},

{'SubModel': 'Camaro SS W/ RS Pkg', 'Trim': 'Camaro SS W/ RS Pkg'},

{'SubModel': '2dr Coupe SS w/2SS', 'Trim': '2dr Coupe SS w/2SS'},

{'SubModel': '2dr Convertible LT w/1LT', 'Trim': '2dr Convertible LT w/1LT'},

{'SubModel': 'N/A', 'Trim': '2LT'},

{'SubModel': "LT RS 6-SPD SUNROOF REAR CAM 20'S", 'Trim': '1LT Coupe 2-Door'},

{'SubModel': '2dr Convertible SS w/2SS', 'Trim': '2dr Convertible SS w/2SS'},

{'SubModel': '2dr Convertible LT w/2LT', 'Trim': '2dr Convertible LT w/2LT'},

{'SubModel': 'N/A', 'Trim': '2LT'},

{'SubModel': 'N/A', 'Trim': 'RARE ZL1 - LOW MILES'},

{'SubModel': "2SS AUTO LEATHER NAV HUD 20'S", 'Trim': 'SS Coupe 2-Door'},

{'SubModel': 'SS', 'Trim': 'SS Coupe 2-Door'},

{'SubModel': 'N/A', 'Trim': 'Car'},

{'SubModel': 'N/A', 'Trim': '2LT'}]

# load data into dataframe

df = pd.DataFrame(data)

# create a dict of all models, including alternative spellings

models = {'LE' : 'LE',

'1LE' : '1LE',

'2LE' : '2LE',

'LT' : 'LT',

'1LT' : '1LT',

'2LT' : '2LT',

'LS' : 'LS',

'1LS' : '1LS',

'2LS' : '2LS',

'SS' : 'SS',

'1SS' : '1SS',

'2SS' : '2SS',

'ZL1' : 'ZL1',

'ZL/1' : 'ZL1',

'ZL-1' : 'ZL1',

'COPO' : 'COPO',

'copo' : 'copo'}

# look for each key in the models dict, and if found, return the value for that key for the column 'TRIM'

def trim_level(row):

for key in models.keys():

if key in (row['Trim'] or row['SubModel']):

return models[key]

df['TRIM'] = df.apply(lambda row: trim_level(row), axis=1)

如下图所示，我现有的解决方案存在一个问题，即2SS被归类为SS，或者2LT被归类为LT。我也不知道如何处理描述中包含两个不同型号的人，例如SS w/2SS。在

^{pr2}$