pandas 將唯一列值轉換為列名，並將所有相關的服務器名稱放在其下

Question

我有下面的代碼，我希望根據ENC1001和ENC1002將服務器名稱放在下面

這些只是兩個ENC1001和ENC1002 ，但我有數百個。

Pandas 代碼：

#!/usr/bin/python3
import pandas as pd
df = pd.read_excel("C7000_Report_Servers_Report.xlsx",  sheet_name=0, usecols=[0, 1, 2, 3])
df = df[df['Enclosure Hardware'].str.contains('C7000')]
print(df)

Output：

          Server Enclosure Hardware      Bay              Server Name
0        ENC1001              C7000    bay 1     dpcfl1001.example.com
1        ENC1001              C7000    bay 2     dpcfl1002.example.com
2        ENC1001              C7000    bay 3     dpcfl1003.example.com
3        ENC1001              C7000    bay 4     dpcfl1004.example.com
4        ENC1001              C7000    bay 5     dpcfl1005.example.com
5        ENC1001              C7000    bay 6     dpcfl1006.example.com
6        ENC1001              C7000    bay 7     dpcfl1007.example.com
7        ENC1001              C7000    bay 8     dpcfl1008.example.com
8        ENC1001              C7000    bay 9     dpcfl1009.example.com
9        ENC1001              C7000   bay 10     dpcfl1010.example.com
10       ENC1001              C7000   bay 11     dpcfl1011.example.com
11       ENC1001              C7000   bay 12                   inc1001
12       ENC1001              C7000   bay 13                   inc1003
13       ENC1001              C7000   bay 14     dpcfl2313.example.com
14       ENC1001              C7000   bay 15                   lic1002
15       ENC1002              C7000    bay 1     dpcfl1012.example.com
16       ENC1002              C7000    bay 2     dpcfl1013.example.com
17       ENC1002              C7000    bay 3     dpcfl1014.example.com
18       ENC1002              C7000    bay 4     dpcfl1015.example.com
19       ENC1002              C7000    bay 5     dpcfl1016.example.com
20       ENC1002              C7000    bay 6     dpcfl1017.example.com
21       ENC1002              C7000    bay 7     dpcfl1018.example.com

想要什么：

ENC1001                             ENC1002                     
dpcfl1001.example.com               dpcfl1012.example.com       
dpcfl1002.example.com               dpcfl1013.example.com       
dpcfl1003.example.com               dpcfl1014.example.com       
dpcfl1004.example.com               dpcfl1015.example.com       
dpcfl1005.example.com               dpcfl1016.example.com       
dpcfl1006.example.com               dpcfl1017.example.com       
dpcfl1007.example.com               dpcfl1018.example.com       
dpcfl1008.example.com               None                           
dpcfl1009.example.com               None                           
dpcfl1010.example.com               None                        
dpcfl1011.example.com               None                            
              inc1001               None                            
              inc1003               None                        
dpcfl2313.example.com               None                            
              lic1002               None                            

謝謝您的幫助。

Answer 1

print(df.head(9))


Server Enclosure Hardware   Bay            Server Name
    0  ENC1001              C7000  bay1  dpcfl1001.example.com
    1  ENC1001              C7000  bay2  dpcfl1002.example.com
    2  ENC1001              C7000  bay3  dpcfl1003.example.com
    3  ENC1001              C7000  bay4  dpcfl1004.example.com
    4  ENC1001              C7000  bay5  dpcfl1005.example.com
    5  ENC1001              C7000  bay6  dpcfl1006.example.com
    6  ENC1001              C7000  bay7  dpcfl1007.example.com
    7  ENC1001              C7000  bay8  dpcfl1008.example.com
    8  ENC1001              C7000  bay9  dpcfl1009.example.com  

 

 

    df2=df.groupby(['Server','Enclosure Hardware','Bay'])['Server Name'].apply(lambda x: pd.Series(x.tolist())).unstack('Server').fillna('None').reset_index().drop(columns=['level_2'])



Server Enclosure Hardware    Bay                ENC1001                ENC1002
0                   C7000   bay1  dpcfl1001.example.com  dpcfl1012.example.com
1                   C7000  bay10  dpcfl1010.example.com                   None
2                   C7000  bay11  dpcfl1011.example.com                   None
3                   C7000  bay12                inc1001                   None
4                   C7000  bay13                inc1003                   None
5                   C7000  bay14  dpcfl2313.example.com                   None
6                   C7000  bay15                lic1002                   None
7                   C7000   bay2  dpcfl1002.example.com  dpcfl1013.example.com
8                   C7000   bay3  dpcfl1003.example.com  dpcfl1014.example.com
9                   C7000   bay4  dpcfl1004.example.com  dpcfl1015.example.com
10                  C7000   bay5  dpcfl1005.example.com  dpcfl1016.example.com
11                  C7000   bay6  dpcfl1006.example.com  dpcfl1017.example.com
12                  C7000   bay7  dpcfl1007.example.com  dpcfl1018.example.com
13                  C7000   bay8  dpcfl1008.example.com                   None
14                  C7000   bay9  dpcfl1009.example.com                   None

Answer 2

在數據框中，同一行的不同列的不同值之間存在一些關系。 所以，我認為使用數據框來獲得你想要的並不是最好的方法。

我建議使用一個字典，其中鍵是Server ，值是Server Name ，如下所示：

from collections import defaultdict

dd = defaultdict(list)
for name, group in df.groupby("Server"):
    dd[name] = group["Server Name"].values

現在，您可以像這樣打印Server的所有Server Name ：

>>> dd["ENC1001"]
['dpcfl1001.example.com' 'dpcfl1002.example.com' 'dpcfl1003.example.com'
 'dpcfl1004.example.com' 'dpcfl1005.example.com' 'dpcfl1006.example.com'
 'dpcfl1007.example.com' 'dpcfl1008.example.com' 'dpcfl1009.example.com'
 'dpcfl1010.example.com' 'dpcfl1011.example.com' 'inc1001' 'inc1003'
 'dpcfl2313.example.com' 'lic1002']

>>> dd["ENC1002"]
['dpcfl1012.example.com' 'dpcfl1013.example.com' 'dpcfl1014.example.com'
 'dpcfl1015.example.com' 'dpcfl1016.example.com' 'dpcfl1017.example.com'
 'dpcfl1018.example.com']

如果你想把它轉換成pandas.DataFrame ，你可以簡單地運行：

>>> new_df = pd.DataFrame.from_dict(dd, orient='index').T
                  ENC1001                ENC1002
0   dpcfl1001.example.com  dpcfl1012.example.com
1   dpcfl1002.example.com  dpcfl1013.example.com
2   dpcfl1003.example.com  dpcfl1014.example.com
3   dpcfl1004.example.com  dpcfl1015.example.com
4   dpcfl1005.example.com  dpcfl1016.example.com
5   dpcfl1006.example.com  dpcfl1017.example.com
6   dpcfl1007.example.com  dpcfl1018.example.com
7   dpcfl1008.example.com                   None
8   dpcfl1009.example.com                   None
9   dpcfl1010.example.com                   None
10  dpcfl1011.example.com                   None
11                inc1001                   None
12                inc1003                   None
13  dpcfl2313.example.com                   None
14                lic1002                   None

Answer 3

你可以這樣做：

pd.concat([g.set_index('Bay').add_suffix(f'_{n}') for n, g in df.groupby('Server')], 
          axis=1, sort=False).filter(like='Server Name')

Output：

          Server Name_ENC1001    Server Name_ENC1002
bay 1   dpcfl1001.example.com  dpcfl1012.example.com
bay 2   dpcfl1002.example.com  dpcfl1013.example.com
bay 3   dpcfl1003.example.com  dpcfl1014.example.com
bay 4   dpcfl1004.example.com  dpcfl1015.example.com
bay 5   dpcfl1005.example.com  dpcfl1016.example.com
bay 6   dpcfl1006.example.com  dpcfl1017.example.com
bay 7   dpcfl1007.example.com  dpcfl1018.example.com
bay 8   dpcfl1008.example.com                    NaN
bay 9   dpcfl1009.example.com                    NaN
bay 10  dpcfl1010.example.com                    NaN
bay 11  dpcfl1011.example.com                    NaN
bay 12                inc1001                    NaN
bay 13                inc1003                    NaN
bay 14  dpcfl2313.example.com                    NaN
bay 15                lic1002                    NaN