Pandas to_json()

import pandas as pd

data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
    'City': ['New York', 'San Francisco', 'Los Angeles']
}

df = pd.DataFrame(data)

# write df to a json file
df.to_json('output.json')

'''
output.json

{"Name":{"0":"Alice","1":"Bob","2":"Charlie"},"Age":{"0":25,"1":30,"2":35},"City":{"0":"New York","1":"San Francisco","2":"Los Angeles"}}
'''

to_json() Syntax

The syntax of the to_json() method in Pandas is:

df.to_json(path_or_buf=None, orient=None, date_format=None, double_precision=10, force_ascii=True, date_unit='ms', default_handler=None, lines=False, compression='infer', index=True, indent=None)

to_json() Arguments

The to_json() method takes the following common arguments:

• path_or_buf (optional): the file path or object where the JSON will be saved
• orient (optional): the format of the JSON string
• date_format (optional): controls the formatting of date values
• double_precision (optional): specifies the number of decimal places for floating-point numbers
• force_ascii (optional): forces ASCII encoding for strings
• date_unit (optional): the time unit to encode for datetime values
• default_handler (optional): function to be called for objects that can't otherwise be serialized
• lines (optional): whether to write the file as a JSON object per line
• compression (optional): specifies the compression format for the output file
• index (optional): include the index as part of the JSON string or file
• indent (optional): Sets the indentation level for pretty-printed JSON output

to_json() Return Value

The return value of to_json() is either None, when writing to a file, or the JSON-formatted string representation of the DataFrame, when no file path is specified.

Example 1: Write to a JSON File

import pandas as pd

# create a DataFrame
data = {'Name': ['Tom', 'Nick', 'John'],
        'Age': [20, 21, 19],
        'City': ['New York', 'London', 'Paris']}
df = pd.DataFrame(data)

# write to JSON file
df.to_json(path_or_buf='output.json')

output.json

{"Name":{"0":"Tom","1":"Nick","2":"John"},"Age":{"0":20,"1":21,"2":19},"City":{"0":"New York","1":"London","2":"Paris"}}

In this example, we wrote df to the JSON file output.json.

Example 2: Using Different Orientations in JSON

import pandas as pd

# create DataFrame
data = {'Name': ['Tom', 'Nick', 'John'],
        'Age': [20, 21, 19],
        'City': ['New York', 'London', 'Paris']}
df = pd.DataFrame(data)

# save DataFrame in different JSON orientations
df.to_json('output_columns.json', orient='columns')
df.to_json('output_records.json', orient='records')

output_columns.json

{"Name":{"0":"Tom","1":"Nick","2":"John"},"Age":{"0":20,"1":21,"2":19},"City":{"0":"New York","1":"London","2":"Paris"}}

output_records.json

[{"Name":"Tom","Age":20,"City":"New York"},{"Name":"Nick","Age":21,"City":"London"},{"Name":"John","Age":19,"City":"Paris"}]

In this example, we saved the DataFrame with different orientations: columns which is the default, and records which gives a list of records.

Notice the difference between the contents of the two JSON files.

Example 3: Writing JSON with Different Date Formats

import pandas as pd
from datetime import datetime

# create DataFrame
data = {'Name': ['Tom', 'Nick', 'John'],
        'Birthdate': [datetime(1998, 5, 12), datetime(1997, 8, 15), datetime(1996, 12, 25)]}
df = pd.DataFrame(data)

# save DataFrame with different date formats
df.to_json('output_epoch.json', date_format='epoch')
df.to_json('output_iso.json', date_format='iso')

output_epoch.json

{"Name":{"0":"Tom","1":"Nick","2":"John"},"Birthdate":{"0":894931200000,"1":871603200000,"2":851472000000}}

output_iso.json

{"Name":{"0":"Tom","1":"Nick","2":"John"},"Birthdate":{"0":"1998-05-12T00:00:00.000","1":"1997-08-15T00:00:00.000","2":"1996-12-25T00:00:00.000"}}

In this example, we formatted date values in JSON output using different date_format options.

Here,

• epoch: represents the number of seconds that have elapsed since January 1, 1970, at 00:00:00 UTC
• iso: is an international standard for representing dates and times

Example 4: Pretty Print JSON Output

import pandas as pd

# create DataFrame
data = {'Name': ['Tom', 'Nick', 'John'],
        'Age': [20, 21, 19]}
df = pd.DataFrame(data)

# pretty print JSON to string
json_str = df.to_json(indent=4)
print(json_str)

Output

{
    "Name":{
        "0":"Tom",
        "1":"Nick",
        "2":"John"
    },
    "Age":{
        "0":20,
        "1":21,
        "2":19
    }
}

In this example, we used indent=4 for pretty-printing the JSON output.

Example 5: Writing JSON with Non-ASCII Characters

import pandas as pd

# create DataFrame with non-ASCII characters
data = {'Name': ['Tomás', 'Niño', 'Jöhn'],
        'Age': [20, 21, 19]}
df = pd.DataFrame(data)

# cave DataFrame with non-ASCII characters
df.to_json('output_ascii.json')
df.to_json('output_non_ascii.json', force_ascii=False)

output_ascii.json

{"Name":{"0":"Tom\u00e1s","1":"Ni\u00f1o","2":"J\u00f6hn"},"Age":{"0":20,"1":21,"2":19}}

output_non_ascii.json

{"Name":{"0":"Tomás","1":"Niño","2":"Jöhn"},"Age":{"0":20,"1":21,"2":19}}

Here, we used force_ascii=False to allow non-ASCII characters like ñ, ö etc. in the JSON file.