在 python pandas 中从相同 csv（我自己的 json 格式）中的值创建一个 Json 列

Posted 2023-03-12

【中文标题】在 python pandas 中从相同 csv（我自己的 json 格式）中的值创建一个 Json 列

【英文标题】：Create a Json column from values in the same csv (My own json format) in python pandas

【发布时间】：2020-07-26 03:41:57

【问题描述】：

我有一个数据框-

+----------+----------+--------+
|     FNAME|     LNAME|     AGE| JSON FOR EACH CUSTOMER
+----------+----------+--------+
|      EARL|     JONES|      35|   ? 
|      MARK|      WOOD|      20|   ?
+----------+----------+--------+

而且我有一个 json 格式 -

  "applicantSegment": 
    "applicantName": 
      "name1": "" ,
      "name3": "" 
    ,
    "ids": [
      
        "type": "",
        "value": ""
      
    ],
    "phones": [
      
        "type": "",
        "value":""  
      
    ],
    "emails": [
      
        "value": "" 
      
    ],
    "dob": 
      "dobDate": "" ,
      "age" : ""  
    
  

我需要在数据框中创建一个新的 json 列，以便新列的每一行都有 json 值。例如，第一行的新列值将是 -

  "applicantSegment": 
    "applicantName": 
      "name1": "EARL" ,
      "name3": "JONES" 
    ,
    "ids": [
      
        "type": "",
        "value": ""
      
    ],
    "phones": [
      
        "type": "",
        "value":""  
      
    ],
    "emails": [
      
        "value": "" 
      
    ],
    "dob": 
      "dobDate": "" ,
      "age": 35  
    
  

有人可以帮忙吗？ 具有新 json 行的必需数据框

+----------+----------+--------+--------------------------------------------------------------
|     FNAME|     LNAME|     AGE| JSON FOR EACH CUSTOMER
+----------+----------+--------+--------------------------------------------------------------
|      EARL|     JONES|      35| "applicantSegment": "applicantName"name1":
|          |                   |  "EARL","name3":"JONES" ,"ids": ["type": "","value": ""],
|          |                   |  "phones""type":"","value":"" ],"emails": ["value": "" ], 
|          |                   |   "dob": "dobDate": "" ,"age":35  
|----------|----------|--------| ---------------------------------------------------------------                                                                             
|          |                   |  Same as above but values from this row
|      MARK|      WOOD|      20|                                                

【参考方案1】：

这是一个 hack，但我之前回答错了你，感觉很糟糕，所以我想我会给你一些至少像你描述的那样有效的东西，直到其他人给出更好的答案。

import pandas as pd
import json

df = pd.DataFrame('FNAME': ['EARL','MARK'],
                  'LNAME': ['JONES','WOOD'],
                  'AGE': ["35","20"])

df['json_representation'] = """ "applicantSegment":  "applicantName":  "name1": """ + df['FNAME'] + """, "name3": """ + df['LNAME'][i] + """  , "ids": [  "type": "", "value": ""  ], "phones": [  "type": "", "value":""   ], "emails": [  "value": ""   ], "dob":  "dobDate": "" , "age" : """ + df['AGE'][i] + """    )"""

我认为真正的解决方案是您应该使用 python 字典来构建它，但我没有看到您有任何可以用作键的独特项。

d = "applicantSegment":  "applicantName":  "name1": "" , "name3": ""  , "ids": [  "type": "", "value": ""  ], "phones": [  "type": "", "value":""   ], "emails": [  "value": ""   ], "dob":  "dobDate": "" , "age" : ""    

s = ""
for i in df.index:
    d.update( "applicantSegment":  "applicantName":  "name1": df['FNAME'][i] , "name3": df['LNAME'][i]  , "ids": [  "type": "", "value": ""  ], "phones": [  "type": "", "value":""   ], "emails": [  "value": ""   ], "dob":  "dobDate": "" , "age" : df['AGE'][i]    )
    s = s + ', ' + json.dumps(d)

s

【讨论】：

嗨，乔希，有没有办法在没有 for 循环的情况下解决这个问题，我有大约 50 万条记录？