pandas if多条件组合场景下精确匹配失效如何解决

你的代码逻辑存在关联校验缺失的问题：两个过滤条件是独立判断的，df.first_unlist.str.match(pat=l.split(' ', 1)[0]).any() 仅判断df中存在任意行匹配当前句子首词，df.Second.str.match('noun').any() 仅判断整个df中存在任意行Second列值为noun，并没有要求匹配首词的那一行对应的Second列值为noun，因此只要df中本身存在Second为noun的行，所有首词匹配的句子都会被保留，导致过滤失效。

你可以将两个条件合并为行筛选逻辑，同时判断同一行是否满足首词匹配、Second列值为noun，另外可将first_unlist的生成移到循环外避免重复计算，提升运行效率：

import pandas as pd

# 原始DataFrame定义
data = {'First':  [['First', 'value'],['second','value'],['third','value','is'],['fourth','value','is']],
'Second': ['noun','not noun','noun', 'not noun']}
df = pd.DataFrame (data, columns = ['First','Second'])

data2 = {'example':  ['First value is important', 'second value is important too','it us good to know',
                  'Firstap is also good', 'aplsecond is very good']}
df2 = pd.DataFrame (data2, columns = ['example'])

# 预先生成拼接后的First列字段，无需重复计算
df['first_unlist'] = [','.join(map(str, l)) for l in df.First]

def checker():
    result =[]
    for sent in df2.example:
        # 提取当前句子首词
        first_word = sent.split(' ', 1)[0]
        # 筛选同时满足首词匹配、Second列为noun的行，判断是否存在
        has_match = (df.first_unlist.str.match(pat=first_word) & (df.Second == 'noun')).any()
        if has_match:
            result.append(sent)
    return result

# 输出结果：['First value is important']
print(checker())

如果数据量较大，你可以提前把所有Second为noun的行对应的首词提取为集合，循环时直接做集合成员判断即可，无需每次遍历整个DataFrame：

# 提前提取符合条件的首词集合
noun_head_words = set(row['First'][0] for _, row in df[df.Second == 'noun'].iterrows())

def checker_fast():
    result = []
    for sent in df2.example:
        sent_head = sent.split(' ', 1)[0]
        if sent_head in noun_head_words:
            result.append(sent)
    return result

内容的提问来源于stack exchange，提问作者zara kolagar