如何在Pandas的逗号分隔列中用REGEX移除[CS][LV]前的文本及括号内容？

需求：删除DataFrame某列中所有包含[CS]、[LV]的片段（包括标签本身及其前面的文本），处理后的列保留剩余的逗号分隔内容。

输入数据 #

colA
My Company Ltd [CS], address, nbc [LV], state [NP], pc [SS], country
Business Plc [CS], address, abc [LV], state [NP], code [SS], country
Work Harder Inc [CS], address, xyz[CS], state [NP], code [SS], country
Company Business People [CS], address, typode [SS], country, nlp [CS]

解决方案 #

正则表达式说明 #

使用正则r'[^,]*\[(CS|LV)\],?\s*'匹配需要删除的片段，各部分含义：

• [^,]*：匹配逗号前的任意非逗号字符（即标签前的所有文本）
• \[(CS|LV)\]：精确匹配[CS]或[LV]，转义方括号是因为其在正则中属于特殊语法
• ,?\s*：匹配可选的逗号及后续任意空白字符，避免删除后留下多余的分隔符或空格

完整代码实现 #

import pandas as pd

# 构造示例DataFrame
data = {
    'colA': [
        'My Company Ltd [CS], address, nbc [LV], state [NP], pc [SS], country',
        'Business Plc [CS], address, abc [LV], state [NP], code [SS], country',
        'Work Harder Inc [CS], address, xyz[CS], state [NP], code [SS], country',
        'Company Business People [CS], address, typode [SS], country, nlp [CS]'
    ]
}
df = pd.DataFrame(data)

# 执行正则替换
df['colA'] = df['colA'].str.replace(r'[^,]*\[(CS|LV)\],?\s*', '', regex=True)

# 可选：处理可能存在的首尾多余逗号/空格（根据实际情况）
df['colA'] = df['colA'].str.strip(', ')

print(df)

处理结果 #

colA
address, state [NP], pc [SS], country
address, state [NP], code [SS], country
address, state [NP], code [SS], country
address, typode [SS], country

内容的提问来源于stack exchange，提问作者Santoo