PROSAGA码农传奇-智慧城市-通过合并列名和其他列值来创建新列

<div class =“post-text”itemprop =“text”>
  
    我删除了额外的列，只关注必要的列进行演示。
  
  
    输入：
  
  
    的
      DF1
    </强>
  
   <pre>
 <code>
 Home 2012-13 2013-14 2014-15 2015-16 2016-17
0 Cleveland Cavaliers 1 1 2 1 3
1 Los Angeles Lakers 2 1 1 1 0
2 Miami Heat 3 3 2 2 1
3 Chicago Bulls 2 1 2 2 1
4 Detroit Pistons 0 0 0 1 1
5 Los Angeles Clippers 2 2 2 1 1
6 New Orleans Pelicans 0 1 1 1 1
7 Philadelphia 76ers 1 0 0 0 0
8 Phoenix Suns 0 0 0 0 0
9 Portland Trail Blazers 1 2 2 0 0
10 Toronto Raptors 0 1 1 2 2

</code>
 </pre>
  
    的
      DF2
    </强>
  
   <pre>
 <code>
 Visitor Home Season
0 Washington Wizards Cleveland Cavaliers 2012-13
1 Dallas Mavericks Los Angeles Lakers 2012-13
2 Boston Celtics Miami Heat 2012-13
3 Dallas Mavericks Utah Jazz 2012-13
4 San Antonio Spurs New Orleans Pelicans 2012-13

</code>
 </pre>
  
    的
      第1步：熔化df1以获取allstars列
    </强>
  
   <pre>
 <code>
 df3 = pd.melt(df1, id_vars='Home', value_vars = df1.columns[df.columns.str.contains('20')], var_name = 'Season', value_name='H_Allstars')

</code>
 </pre>
  
    输出继电器：
  
   <pre>
 <code>
 Home Season H_Allstars
0 Cleveland Cavaliers 2012-13 1
1 Los Angeles Lakers 2012-13 2
2 Miami Heat 2012-13 3
3 Chicago Bulls 2012-13 2
4 Detroit Pistons 2012-13 0
5 Los Angeles Clippers 2012-13 2
6 New Orleans Pelicans 2012-13 0
7 Philadelphia 76ers 2012-13 1
8 Phoenix Suns 2012-13 0
...

</code>
 </pre>
  
    的
      第2步：将此新数据框与df2合并，以获取H_Allstars和V_Allstars列
    </强>
  
   <pre>
 <code>
 df4 = pd.merge(df2, df3, how='left', on=['Home', 'Season'])

</code>
 </pre>
  
    输出：
  
   <pre>
 <code>
 Visitor Home Season H_Allstars
0 Washington Wizards Cleveland Cavaliers 2012-13 1.0
1 Dallas Mavericks Los Angeles Lakers 2012-13 2.0
2 Boston Celtics Miami Heat 2012-13 3.0
3 Dallas Mavericks Utah Jazz 2012-13 NaN
4 San Antonio Spurs New Orleans Pelicans 2012-13 0.0

</code>
 </pre>
  
    的
      第3步：添加V_Allstars列
    </强>
  
   <pre>
 <code>
 # renaming column as required
df3.rename(columns={'Home': 'Visitor', 'H_Allstars': 'V_Allstars'}, inplace=True)

df5 = pd.merge(df4, df3, how='left', on=['Visitor', 'Season'])

</code>
 </pre>
  
    输出：
  
   <pre>
 <code>
 Visitor Home Season H_Allstars V_Allstars
0 Washington Wizards Cleveland Cavaliers 2012-13 1.0 NaN
1 Dallas Mavericks Los Angeles Lakers 2012-13 2.0 NaN
2 Boston Celtics Miami Heat 2012-13 3.0 NaN
3 Dallas Mavericks Utah Jazz 2012-13 NaN NaN
4 San Antonio Spurs New Orleans Pelicans 2012-13 0.0 NaN

</code>
 </pre>
</DIV>