从 pandas 数据帧构造 JSON

Question

I have this panda data frame containg info about id, description, field name, field value and email

                    id                    desc   field_name       val               eml  
0                    1            desc in id 1      field_1       300    test@gmail.com  
1                    1            desc in id 1      field_2       305    test@gmail.com  
2                    2            desc in id 2      field_3       530    test2@gmail.com   
3                    2            desc in id 2      field_4       745    test2@gmail.com   
3                    2            desc in id 2      field_1       845    test2@gmail.com   

Here desc and email will be same for a particular id So I want to construct an array of JSON by clubbing the id, desc and email

And all the changes on field name and val will be send by list of list format

[
    {
        id: 1,
        desc: "desc in id 1",
        eml: "test@gmail.com"
        values: [ [field_1, 300], [field_2, 305]  ]
    }, 
    {
        id: 2,
        desc: "desc in id 2",
        eml: "test2@gmail.com"
        values: [ [field_3, 530], [field_4, 745], [field_1, 845]  ]
    }, 
]

Answer 1

(If you google "pandas to json" the first hit is the link below...)

You can use the pd.to_json ie

df_json = df.to_json()

where the argument orient defines different outputs. You'll most likely use orient="records" to get it as a json-array eg

import pandas as pd
df_so = pd.DataFrame({"id":[1,2,3],"email":["foo@example.com","bar@example.com","baz@example.com"]})
print(df_so)
#   id  email
# 0 1   foo@example.com
# 1 2   bar@example.com
# 2 3   baz@example.com

df_so.to_json(orient="records")

#
#[{"id":1,"email":"foo@example.com"},
#{"id":2,"email":"bar@example.com"},
#{"id":3,"email":"baz@example.com"}]'

regarding your values you'll need to create that column first.