python - 如何根据python中的Where函数获取两列值

Question

问题是:基于 user_id 列，我想获取 rating 和 product_id 列的值。可以有多个条目具有相同的 user_id。我想用 rating 和 product_id 列值获取所有用户记录但是对于用户没有评价的电影应该显示为 Nan 但仍然，应检索 product_id。下表提供了一些数据。

| product_id  | user_id         | user_name                                    | rating |
|-------------|-----------------|----------------------------------------------|--------|
|  B0009XRZ92 | A2JFZLAUG3YFQ7  |  Entropy Babe "EB"                           | 5      |
|  B0009XRZ92 | A22HGAAO8KZ2N3  |  R.   Metzelar                               | 5      |
|  B000067A8B |  A2NJO6YE954DBH |  Lawrance M. Bernabo                         | 4      |
|  B0009XRZ92 |  A3HE4MYMWK4AER |  Rebecca M. Eddy "Foster Mom and   Untbunny" | 5      |
|  B003A3R3ZY | A9A2PR663ED1V   |  Roger D. Goff                               | 5      |
|  B0009XRZ92 | A2MRZDJF90JC1U  |  Suzanne K. Armstrong "Suzy Q"               | 5      |
|  B0009XRZ92 |  A2YNBDT3170PCR |  C.   O'Hern                                 | 5      |
|  B0009XRZ92 |  A10VJ7BDVCPKEZ |  Carol S. Bottom                             | 5      |
|  B0009XRZ92 |  AAAQO894MG80B  |  Paul J. Michko                              | 5      |
|  B00067BBQE | A9A2PR663ED1V   |  Roger D. Goff                               | 5      |
|  B0009XRZ92 | A31S5QUMFR8NH2  |  Dana L. Jordan "Mom of Twins"               | 5      |
|  B0009XRZ92 |  A2DS24DHXUH0GM |  Gaz    Rev(iewer)                           | 4      |
|  B00006AUMZ |  A2NJO6YE954DBH |  Lawrance M. Bernabo                         | 4      |
|  B0009XRZ92 |  A16FRHL2ZC7EUR |  M.   Claytor                                | 5      |
|  B0009XRZ92 | A3AV8R0A62PP1N  |  MARCUSHELBLINZ "mmmacman"                   | 5      |
|  B0009XRZ92 |  A3QN84C38DE9FU |  Gillian M. Kratzer                          | 5      |
|  B0009XRZ92 |  A36MLTLVQFEQYL |  Yossarian "alienated socialist"             | 5      |
|  B00006AUMD |  A2NJO6YE954DBH |  Lawrance M. Bernabo                         | 4      |

What I want to do is:

To take one user_id at a time and display the rating and product_id columns value for that user for all the movies in the table and if the user didn't rate some movies then the record should be displayed with the product_id value and rating as Nan and the whole process should be repeated for all the users.

例如，user_id: A2NJO6YE954DBH 的记录将如下所示:

| product_id | rating |
|------------|--------|
| B000067A8B | 4      |
| B00006AUMD | 4      |
| B00006AUMD | 4      |
| B0009XRZ92 | Nan    |
| B003A3R3ZY | Nan    |
| B00067BBQE | Nan    |
|     .      |   .    |
|     .      |   .    |
|     .      |    .   |

我曾尝试使用 Pandas 库为此编写代码，但无法自拔。这就是我所做的一切，但它没有输出我想要的。

import pandas as pd
df =pd.read_csv('out.csv')
unique_users=df.user_id.unique()
for x, y in enumerate(unique_users):
    print(df[['rating','product_id']].where(df.user_id==y))

请帮帮我..谢谢

Answer 1

如果我没理解错的话，你可以使用pd.pivot_table()这里:

df_new=df.pivot_table(index='user_id',columns='product_id',values='rating').rename_axis(None,1)
print(df_new)


                     B000067A8B     B00006AUMD     B00006AUMZ     B00067BBQE   \
user_id                                                                         
  A10VJ7BDVCPKEZ             NaN            NaN            NaN            NaN   
  A16FRHL2ZC7EUR             NaN            NaN            NaN            NaN   
  A2DS24DHXUH0GM             NaN            NaN            NaN            NaN   
  A2NJO6YE954DBH             4.0            4.0            4.0            NaN   
  A2YNBDT3170PCR             NaN            NaN            NaN            NaN   
  A36MLTLVQFEQYL             NaN            NaN            NaN            NaN   
  A3HE4MYMWK4AER             NaN            NaN            NaN            NaN   
  A3QN84C38DE9FU             NaN            NaN            NaN            NaN   
  AAAQO894MG80B              NaN            NaN            NaN            NaN   
 A22HGAAO8KZ2N3              NaN            NaN            NaN            NaN   
 A2JFZLAUG3YFQ7              NaN            NaN            NaN            NaN   
 A2MRZDJF90JC1U              NaN            NaN            NaN            NaN   
 A31S5QUMFR8NH2              NaN            NaN            NaN            NaN   
 A3AV8R0A62PP1N              NaN            NaN            NaN            NaN   
 A9A2PR663ED1V               NaN            NaN            NaN            5.0   

                     B0009XRZ92     B003A3R3ZY   
user_id                                          
  A10VJ7BDVCPKEZ             5.0            NaN  
  A16FRHL2ZC7EUR             5.0            NaN  
  A2DS24DHXUH0GM             4.0            NaN  
  A2NJO6YE954DBH             NaN            NaN  
  A2YNBDT3170PCR             5.0            NaN  
  A36MLTLVQFEQYL             5.0            NaN  
  A3HE4MYMWK4AER             5.0            NaN  
  A3QN84C38DE9FU             5.0            NaN  
  AAAQO894MG80B              5.0            NaN  
 A22HGAAO8KZ2N3              5.0            NaN  
 A2JFZLAUG3YFQ7              5.0            NaN  
 A2MRZDJF90JC1U              5.0            NaN  
 A31S5QUMFR8NH2              5.0            NaN  
 A3AV8R0A62PP1N              5.0            NaN  
 A9A2PR663ED1V               NaN            5.0