Creating a unique json object from pandas

I have this dataframe

       Year     type        Median_Home_Value
786252  2010    analyzed    11973.000
786253  2011    analyzed    12500.000
786254  2012    analyzed    13325.000
786255  2013    analyzed    14204.000
786256  2014    analyzed    14815.000
786257  2015    analyzed    15393.000
786258  2016    analyzed    15901.000
786259  2017    analyzed    16680.000
786260  2018    analyzed    17497.000
786261  2019    analyzed    18249.000
786262  2020    analyzed    19381.000
786263  2021    analyzed    20292.000
899389  2027    predicted   20718.132
899390  2024    predicted   21225.432
899397  2026    predicted   21103.680
899415  2025    predicted   20779.008
899481  2023    predicted   20941.344

I want to create a json object to represent this dataframe to be like this

{

    median_home_value: [{year: 2010, value: 11973, type: analyzed}, {year: 2011, value: 12500, type: analyzed}, ....]

}

I tried to do something like this

d = {}
d['Median_Home_Value] = test_df[['Year', 'Median_Home_Value', 'type']].to_json()

but it does not give me the expected result, any suggestions are appreciated.

>Solution :

Convert the dataframe to dictionary representation then dump the records into json data:

import json

records = df.rename(columns={'Median_Home_Value': 'value'}).to_dict('records')
json_data = json.dumps({'median_home_value': records})

Result

{
    "median_home_value": [
        {
            "Year": 2010,
            "type": "analyzed",
            "value": 11973.0
        },
        {
            "Year": 2011,
            "type": "analyzed",
            "value": 12500.0
        },
        {
            "Year": 2012,
            "type": "analyzed",
            "value": 13325.0
        },
        ...
}