使用正则表达式 pandas 改变年龄答案

【问题标题】：changing age using regex pandas使用正则表达式 pandas 改变年龄
【发布时间】：2019-12-28 05:36:13
【问题描述】：

    import pandas as pd
    dataframe = pd.DataFrame({'Data' : ['A 90-year-old or 96-year-old and 110-year-old is 90 days  ', 
                                       'For all 82-year-old is the 94-year-old why 28A ', 
                                       'But the fact is 101-year-old 109-year-old cool 100',],
                          'ID': [1,2,3]

                         })
#tried this regex
dataframe['New'] = dataframe['Data'].str.replace(r'\d+(-year-old)', r'>90')

dataframe
    Data                                                      ID    New
0   A 90-year-old or 96-year-old and 110-year-old is 90 days  1 A >90 or >90 and >90 is 90 days
1   For all 82-year-old is the 94-year-old why 28A            2 For all >90 is the >90 why 28A
2   But the fact is 101-year-old 109-year-old cool 100        3 But the fact is >90 >90 cool 100

我正在尝试使用正则表达式来更改所有 90 岁以上的年龄。因此，例如，90-year-old 将更改为 >90。但82-year-old 或任何 90 岁以下的人都不应该这样做。如上所示，我已接近我想要的，但 82-year-old 仍会更改为 >90 但它不应该

如何在这行代码中更改我的正则表达式

   dataframe['New'] = dataframe['Data'].str.replace(r'\d+(-year-old)', r'>90')

以便仅将90-year-old 及以上（例如91-year-old、98-year-old、105-year-old 等）更改为>90？

【问题讨论】：

标签： python regex string pandas replace

【解决方案1】：

您可以使用涵盖两种情况的正则表达式来指定这一点：9[1-9] 和 \d{3,}：

dataframe['New'] = dataframe['Data'].str.replace(r'<b>(9[1-9]|\d{3,})</b>(-year-old)', r'&gt;90')

第一部分9[1-9] 因此匹配91 和99 之间的所有值，第二部分匹配所有三位或更多数字（1234 当然非常不太可能） .

对于给定的样本数据，我们得到：

>>> dataframe['Data'].str.replace(r'(9[1-9]|\d{3,})(-year-old)', r'>90')
0    A 90-year-old or >90 and >90 is 90 days  
1      For all 82-year-old is the >90 why 28A 
2             But the fact is >90 >90 cool 100
Name: Data, dtype: object

如果要包含90，可以将正则表达式更改为：

>>> dataframe['Data'].str.replace(r'(9\d|\d{3,})(-year-old)', r'>90')
0          A >90 or >90 and >90 is 90 days  
1    For all 82-year-old is the >90 why 28A 
2           But the fact is >90 >90 cool 100
Name: Data, dtype: object

【讨论】：