---

Often you might be interested in converting a pandas DataFrame to a JSON format.

Fortunately this is easy to do using the  function, which allows you to convert a DataFrame to a JSON string with one of the following formats:

• ‘split’ : dict like {‘index’ -> [index], ‘columns’ -> [columns], ‘data’ -> [values]}
• ‘records’ : list like [{column -> value}, … , {column -> value}]
• ‘index’ : dict like {index -> {column -> value}}
• ‘columns’ : dict like {column -> {index -> value}}
• ‘values’ : just the values array
• ‘table’ : dict like {‘schema’: {schema}, ‘data’: {data}}

This tutorial shows how to convert a DataFrame to each of the six formats using the following pandas DataFrame:

import pandas as pd

#create DataFrame
df = pd.DataFrame({'points': [25, 12, 15, 19],
                   'assists': [5, 7, 7, 12]})  

#view DataFrame
df

        points	assists
0	25	5
1	12	7
2	15	7
3	19	12

Method 1: ‘Split’

df.to_json(orient='split')

{
   "columns": [
      "points",
      "assists"
   ],
   "index": [
      0,
      1,
      2,
      3
   ],
   "data": [
      [
         25,
         5
      ],
      [
         12,
         7
      ],
      [
         15,
         7
      ],
      [
         19,
         12
      ]
   ]
}

Method 2: ‘Records’

df.to_json(orient='records')

[
   {
      "points": 25,
      "assists": 5
   },
   {
      "points": 12,
      "assists": 7
   },
   {
      "points": 15,
      "assists": 7
   },
   {
      "points": 19,
      "assists": 12
   }
] 

Method 3: ‘Index’

df.to_json(orient='index') 

{
   "0": {
      "points": 25,
      "assists": 5
   },
   "1": {
      "points": 12,
      "assists": 7
   },
   "2": {
      "points": 15,
      "assists": 7
   },
   "3": {
      "points": 19,
      "assists": 12
   }
}

Method 4: ‘Columns’

df.to_json(orient='columns') 

{
   "points": {
      "0": 25,
      "1": 12,
      "2": 15,
      "3": 19
   },
   "assists": {
      "0": 5,
      "1": 7,
      "2": 7,
      "3": 12
   }
}

Method 5: ‘Values’

df.to_json(orient='values') 

[
   [
      25,
      5
   ],
   [
      12,
      7
   ],
   [
      15,
      7
   ],
   [
      19,
      12
   ]
]

Method 6: ‘Table’

df.to_json(orient='table') 

{
   "schema": {
      "fields": [
         {
            "name": "index",
            "type": "integer"
         },
         {
            "name": "points",
            "type": "integer"
         },
         {
            "name": "assists",
            "type": "integer"
         }
      ],
      "primaryKey": [
         "index"
      ],
      "pandas_version": "0.20.0"
   },
   "data": [
      {
         "index": 0,
         "points": 25,
         "assists": 5
      },
      {
         "index": 1,
         "points": 12,
         "assists": 7
      },
      {
         "index": 2,
         "points": 15,
         "assists": 7
      },
      {
         "index": 3,
         "points": 19,
         "assists": 12
      }
   ]
}

How to Export a JSON File

You can use the following syntax to export a JSON file to a specific file path on your computer:

#create JSON file 
json_file = df.to_json(orient='records') 

#export JSON file
with open('my_data.json', 'w') as f:
    f.write(json_file)

You can find the complete documentation for the pandas to_json() function .