熊猫用另一栏的子串制作新的栏。

Question

我试着用另一个专栏的子字符串在熊猫中创建新的专栏。

import pandas as pd
import re

df = {'title':['Apartment 2 roomns, 40 m²', 'House 7 rooms, 183 m²', 'House 4 rooms, 93 m²', 'Apartment 12 rooms, 275 m²']} 

我正试图用regex来捕获组：

df['Name'] = df.title.str.extract(r'(^[a-zA-Z]+)', expand=True) 

这个我得到了一个很好的结果。但我需要一个有房间数量的栏目(没有“房间”这个词)和另一个没有“平方米”大小的栏目。我试过：

df['Rooms'] = df.title.str.replace(r'(^[0-9]+)\s(rooms)', r'\1') #to capture only the first group, which is the number

df['Size'] = df.title.str.replace(r'(^[0-9]+)\s(m²)', r'\1') #to capture only the first group, which is the number

我的产出：

   Name      Rooms                         Size
0  Apartment Apartment 2 roomns, 40 m²     Apartment 2 roomns, 40 m²
1  House     House 7 rooms, 183 m²         House 7 rooms, 183 m²
2  House     House 4 rooms, 93 m²          House 4 rooms, 93 m²
3  Apartment Apartment 12 rooms, 275 m²    Apartment 12 rooms, 275 m²

良好的产出：

   Name      Rooms Size
0  Apartment 2     40
1  House     7     183
2  House     4     93
3  Apartment 12    275

Answer 1

您可以使用

df["Rooms"] = df["title"].str.extract(r'(\d+)\s*room', expand=False)
df['Size'] = df["title"].str.extract(r'(\d+(?:\.\d+)?)\s*m²', expand=False)

输出：

>>> df
                        title Rooms Size
0   Apartment 2 roomns, 40 m²     2   40
1       House 7 rooms, 183 m²     7  183
2        House 4 rooms, 93 m²     4   93
3  Apartment 12 rooms, 275 m²    12  275

(\d+)\s*room正则表达式匹配并捕获到Group 1中的一个或多个数字，然后只匹配零或多个空白空间(\s*)，然后匹配一个room字符串。

(\d+(?:\.\d+)?)\s*m²正则表达式匹配并捕获一个或多个数字，以及一个.和一个或多个数字的可选字符串，然后匹配零或多个空白空间，然后匹配一个m²字符串。

参见regex #1演示和regex #2演示。