In [1]:

    

import pandas as pd
import pprint
import json


    

In [2]:

    

df = pd.DataFrame({'col1': [1, 2, 3], 'col2': ['a', 'x', 'あ']},
                  index=['row1', 'row2', 'row3'])


    

In [3]:

    

print(df)


    

    




      col1 col2
row1     1    a
row2     2    x
row3     3    あ

In [4]:

    

print(df.to_json())


    

    




{"col1":{"row1":1,"row2":2,"row3":3},"col2":{"row1":"a","row2":"x","row3":"\u3042"}}

In [5]:

    

print(type(df.to_json()))


    

    




<class 'str'>

In [6]:

    

path = 'data/src/sample_from_pandas_columns.json'


    

In [7]:

    

df.to_json(path)


    

In [8]:

    

with open(path) as f:
    s = f.read()
    print(s)
    print(type(s))


    

    




{"col1":{"row1":1,"row2":2,"row3":3},"col2":{"row1":"a","row2":"x","row3":"\u3042"}}
<class 'str'>

In [9]:

    

with open(path, encoding='unicode-escape') as f:
    s = f.read()
    print(s)
    print(type(s))


    

    




{"col1":{"row1":1,"row2":2,"row3":3},"col2":{"row1":"a","row2":"x","row3":"あ"}}
<class 'str'>

In [10]:

    

with open(path) as f:
    d = json.load(f)
    print(d)
    print(type(d))


    

    




{'col1': {'row1': 1, 'row2': 2, 'row3': 3}, 'col2': {'row1': 'a', 'row2': 'x', 'row3': 'あ'}}
<class 'dict'>

In [11]:

    

df_read = pd.read_json(path)
print(df_read)


    

    




      col1 col2
row1     1    a
row2     2    x
row3     3    あ

In [12]:

    

df.to_json('data/src/sample_from_pandas_columns.gz', compression='gzip')


    

In [13]:

    

print(df.to_json(orient='split'))


    

    




{"columns":["col1","col2"],"index":["row1","row2","row3"],"data":[[1,"a"],[2,"x"],[3,"\u3042"]]}

In [14]:

    

pprint.pprint(json.loads(df.to_json(orient='split')))


    

    




{'columns': ['col1', 'col2'],
 'data': [[1, 'a'], [2, 'x'], [3, 'あ']],
 'index': ['row1', 'row2', 'row3']}

In [15]:

    

print(df.to_json(orient='records'))


    

    




[{"col1":1,"col2":"a"},{"col1":2,"col2":"x"},{"col1":3,"col2":"\u3042"}]

In [16]:

    

pprint.pprint(json.loads(df.to_json(orient='records')), width=40)


    

    




[{'col1': 1, 'col2': 'a'},
 {'col1': 2, 'col2': 'x'},
 {'col1': 3, 'col2': 'あ'}]

In [17]:

    

print(df.to_json(orient='records', lines=True))


    

    




{"col1":1,"col2":"a"}
{"col1":2,"col2":"x"}
{"col1":3,"col2":"\u3042"}

In [18]:

    

print(df.to_json(orient='index'))


    

    




{"row1":{"col1":1,"col2":"a"},"row2":{"col1":2,"col2":"x"},"row3":{"col1":3,"col2":"\u3042"}}

In [19]:

    

pprint.pprint(json.loads(df.to_json(orient='index')))


    

    




{'row1': {'col1': 1, 'col2': 'a'},
 'row2': {'col1': 2, 'col2': 'x'},
 'row3': {'col1': 3, 'col2': 'あ'}}

In [20]:

    

print(df.to_json(orient='columns'))


    

    




{"col1":{"row1":1,"row2":2,"row3":3},"col2":{"row1":"a","row2":"x","row3":"\u3042"}}

In [21]:

    

pprint.pprint(json.loads(df.to_json(orient='columns')))


    

    




{'col1': {'row1': 1, 'row2': 2, 'row3': 3},
 'col2': {'row1': 'a', 'row2': 'x', 'row3': 'あ'}}

In [22]:

    

print(df.to_json(orient='values'))


    

    




[[1,"a"],[2,"x"],[3,"\u3042"]]

In [23]:

    

pprint.pprint(json.loads(df.to_json(orient='values')))


    

    




[[1, 'a'], [2, 'x'], [3, 'あ']]

In [24]:

    

print(df.to_json(orient='table'))


    

    




{"schema": {"fields":[{"name":"index","type":"string"},{"name":"col1","type":"integer"},{"name":"col2","type":"string"}],"primaryKey":["index"],"pandas_version":"0.20.0"}, "data": [{"index":"row1","col1":1,"col2":"a"},{"index":"row2","col1":2,"col2":"x"},{"index":"row3","col1":3,"col2":"\u3042"}]}

In [25]:

    

pprint.pprint(json.loads(df.to_json(orient='table')))


    

    




{'data': [{'col1': 1, 'col2': 'a', 'index': 'row1'},
          {'col1': 2, 'col2': 'x', 'index': 'row2'},
          {'col1': 3, 'col2': 'あ', 'index': 'row3'}],
 'schema': {'fields': [{'name': 'index', 'type': 'string'},
                       {'name': 'col1', 'type': 'integer'},
                       {'name': 'col2', 'type': 'string'}],
            'pandas_version': '0.20.0',
            'primaryKey': ['index']}}

In [26]:

    

print(df.to_json(force_ascii=False))


    

    




{"col1":{"row1":1,"row2":2,"row3":3},"col2":{"row1":"a","row2":"x","row3":"あ"}}