在 Pandas DataFrame 列中将单列拆分为多列

Luqman Khan 2022年5月16日
在 Pandas DataFrame 列中将单列拆分为多列

Pandas 有一种众所周知的方法,可以通过列表的破折号、空格和返回列(Series)来拆分字符串列或文本列;如果我们谈论 pandas,术语 Series 被称为 Dataframe 列。

我们可以使用 pandas Series.str.split() 函数将字符串拆分为围绕给定分隔符或定界符的多列。它类似于 Python 字符串 split() 方法,但适用于整个 Dataframe 列。我们有最简单的方法来分隔下面的列。

此方法将 Series 字符串与初始索引分开。

Series.str.split(pat=None, n=-1, expand=False)

让我们尝试了解此方法的工作原理

# import Pandas as pd
import pandas as pd

# innitilize Dataframe
df = pd.DataFrame(
    {
        "Email": [
            "Alex.jhon@gmail.com",
            "Hamza.Azeez@gmail.com",
            "Harry.barton@hotmail.com",
        ],
        "Number": ["+44-3844556210", "+44-2245551219", "+44-1049956215"],
        "Location": ["Alameda,California", "Sanford,Florida", "Columbus,Georgia"],
    }
)
print("Dataframe series:\n", df)

我们创建了一个 Dataframe df,包含三列,EmailNumberLocation。请注意,电子邮件列中的字符串具有特定的模式。但是,如果你仔细观察,可以将此列拆分为两列。我们将很好地解决所需的问题。

输出:

Dataframe series :
                       Email          Number            Location
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia

我们将使用 Series.str.split() 函数来分隔 Number 列并在 split() 方法中传递 -。确保将 True 传递给 expand 关键字。

示例 1:

print(
    "\n\nSplit 'Number' column by '-' into two individual columns :\n",
    df.Number.str.split(pat="-", expand=True),
)

这个例子将用 - 分割系列(数字)的每个值。

输出:

Split 'Number' column into two individual columns :
      0           1
0  +44  3844556210
1  +44  2245551219
2  +44  1049956215

如果我们只使用扩展参数 Series.str.split(expand=True),这将允许拆分空格,但不能用 -, 或字符串中存在的任何正则表达式进行分隔,你必须通过 pat 参数。

让我们重命名这些拆分列。

df[["Dialling Code", "Cell-Number"]] = df.Number.str.split("-", expand=True)
print(df)

我们创建了两个新系列 Dialling codeCell-Number 并使用 Number 系列分配值。

输出:

                      Email          Number            Location Dialling Code  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California           +44   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida           +44   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia           +44   

  Cell-Number  
0  3844556210  
1  2245551219  
2  1049956215  

示例 2:

在这个例子中,我们将用 , 分割 Location 系列。

df[["City", "State"]] = df.Location.str.split(",", expand=True)
print(df)

拆分 Location 系列并将其值存储在单独的系列 CityState 中。

输出:

                      Email          Number            Location      City  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California   Alameda   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida   Sanford   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia  Columbus   

        State  
0  California  
1     Florida  
2     Georgia 

让我们看看最后一个例子。我们将在 Email 系列中分隔全名。

full_name = df.Email.str.split(pat="@", expand=True)
print(full_name)

输出:

              0            1
0     Alex.jhon    gmail.com
1   Hamza.Azeez    gmail.com
2  Harry.barton  hotmail.com

现在我们用 . 分隔名字和姓氏。

df[["First Name", "Last Name"]] = full_name[0].str.split(".", expand=True)
print(df)

输出:

                      Email          Number            Location First Name  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California       Alex   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida      Hamza   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia      Harry   

  Last Name  
0      jhon  
1     Azeez  
2    barton 

如果在 .split() 方法中传递了 expand=Truen=-1 参数将不起作用。

print(df["Email"].str.split("@", n=-1, expand=True))

输出:

        0           1
0  George  Washington
1   Hamza       Azeez
2   Harry      Walker

整个示例代码如下。

# import Pandas as pd
import pandas as pd

# create a new Dataframe
df = pd.DataFrame(
    {
        "Email": [
            "Alex.jhon@gmail.com",
            "Hamza.Azeez@gmail.com",
            "Harry.barton@hotmail.com",
        ],
        "Number": ["+44-3844556210", "+44-2245551219", "+44-1049956215"],
        "Location": ["Alameda,California", "Sanford,Florida", "Columbus,Georgia"],
    }
)

print("Dataframe series :\n", df)

print(
    "\n\nSplit 'Number' column by '-' into two individual columns :\n",
    df.Number.str.split(pat="-", expand=True),
)

df[["Dialling Code", "Cell-Number"]] = df.Number.str.split("-", expand=True)
print(df)

df[["City", "State"]] = df.Location.str.split(",", expand=True)
print(df)

full_name = df.Email.str.split(pat="@", expand=True)
print(full_name)

df[["First Name", "Last Name"]] = full_name[0].str.split(".", expand=True)
print(df)

输出:

Dataframe series :
                       Email          Number            Location
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia


Split 'Number' column by '-' into two individual columns :
      0           1
0  +44  3844556210
1  +44  2245551219
2  +44  1049956215
                      Email          Number            Location Dialling Code  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California           +44   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida           +44   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia           +44   

  Cell-Number  
0  3844556210  
1  2245551219  
2  1049956215  
                      Email          Number            Location Dialling Code  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California           +44   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida           +44   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia           +44   

  Cell-Number      City       State  
0  3844556210   Alameda  California  
1  2245551219   Sanford     Florida  
2  1049956215  Columbus     Georgia  
              0            1
0     Alex.jhon    gmail.com
1   Hamza.Azeez    gmail.com
2  Harry.barton  hotmail.com
                      Email          Number            Location Dialling Code  \
0       Alex.jhon@gmail.com  +44-3844556210  Alameda,California           +44   
1     Hamza.Azeez@gmail.com  +44-2245551219     Sanford,Florida           +44   
2  Harry.barton@hotmail.com  +44-1049956215    Columbus,Georgia           +44   

  Cell-Number      City       State First Name Last Name  
0  3844556210   Alameda  California       Alex      jhon  
1  2245551219   Sanford     Florida      Hamza     Azeez  
2  1049956215  Columbus     Georgia      Harry    barton 

相关文章 - Pandas DataFrame