Как преобразовать кадр данных в файл JSON, используя вложенную структуру в повторной группе ⇐ Python

[Anonymous]

Моя команда попросила меня преобразовать кадр данных в файл JSON с определенной вложенной структурой. Мой кадр данных является следующим: < /p>
df = pd.DataFrame({
'type': ['customer','customer','customer','customer','customer','customer','customer','customer','customer','customer','customer','customer','customer','customer','customer'],
'customer_id': ['1-0000001','1-0000001','1-0000001','1-0000001','1-0000002','1-0000002','1-0000002','1-0000002','1-0000002','1-0000002','1-0000003','1-0000003','1-0000003','1-0000003','1-0000003'],
'email': ['customer1@otenet.gr','customer1@otenet.gr','customer1@otenet.gr','customer1@otenet.gr','customer2@gmail.com','customer2@gmail.com','customer2@gmail.com','customer2@gmail.com','customer2@gmail.com','customer2@gmail.com','customer3@yahoo.com.au','customer3@yahoo.com.au','customer3@yahoo.com.au','customer3@yahoo.com.au','customer3@yahoo.com.au'],
'# of policies':[4,4,4,4,6,6,6,6,6,6,5,5,5,5,5],
'POLICY_NO': ['000000001','000000002','000000003','000000004','000000005','000000006','000000007','000000008','000000009','000000010','000000011','000000012','000000013','000000014','000000015'],
'RECEIPT_NO': [420000001,'420000002','420000003','420000004','420000005','420000006','420000007','420000008','420000009','420000010','420000011','420000012','420000013','420000014','420000015'],
'PAYMENT_CODE': ['RF35000000000000000000001','RF35000000000000000000002','RF35000000000000000000003','RF35000000000000000000004','RF35000000000000000000005','RF35000000000000000000006','null','RF35000000000000000000008','RF35000000000000000000009','null','RF35000000000000000000011','RF35000000000000000000012','null','RF35000000000000000000014','RF35000000000000000000015'],
'KLADOS': ['Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου','Αυτοκινήτου']
})

What I would like to achieve is creating a JSON file grouped by the column 'type', 'customer_id' and a dictionary that contains the fields 'email' and '# of policies'.
These three elements are used for grouping.
The attribute dictionary must contain another nested dictionary called 'policy details' with POLICY_NO, RECEIPT_NO, PAYMENT_CODE and Klados.
«Словарь политики» должен повторяться для каждой комбинации «тип», «customer_id» и «атрибуты» следующим образом: < /p>
"type": "customer",
"customer_id": "1-0000001",
"attributes": {
"email": "customer1@otenet.gr",
"# of policies": 4,
"policies details": [
{
"POLICY_NO": 000000001,
"RECEIPT_NO": 420000001,
"PAYMENT_CODE": "RF35000000000000000000001"
"KLADOS": "Αυτοκινήτου"
},
{
"POLICY_NO": 000000002,
"RECEIPT_NO": 420000002,
"PAYMENT_CODE": "RF35000000000000000000002"
"KLADOS": "Αυτοκινήτου"
......
< /code>
Для достижения этого формата я попробовал следующие способы: < /p>

1 -й метод < /p>
# Create the new 'policy_details' field as dictionary
df['policy_details'] = df[['POLICY_NO', 'RECEIPT_NO', 'PAYMENT_CODE', 'KLADOS']].to_dict(orient='records')

# Create the new 'attributes' field as dictionary with the nested 'policy_details'
df['attributes'] = df[['email', '# of policies', 'policy_details']].to_dict(orient='records')

# JSON Section

# Create the JSON structure
j = df[['type', 'customer_id', 'attributes']].to_json(orient='records', force_ascii = False)

print(j)
< /code>
, но < /p>

'type', 'customer_id' и атрибуты не сгруппированы (они повторяются для каждой политики class = "lang-json prettyprint-override">[{
"type": "customer",
"customer_id": "1-0000001",
"attributes": {
"email": "customer1@otenet.gr",
"# of policies": 4,
"policy_details": {
"POLICY_NO": "000000001",
"RECEIPT_NO": 420000001,
"PAYMENT_CODE": "RF35000000000000000000001",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000001",
"attributes": {
"email": "customer1@otenet.gr",
"# of policies": 4,
"policy_details": {
"POLICY_NO": "000000002",
"RECEIPT_NO": "420000002",
"PAYMENT_CODE": "RF35000000000000000000002",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000001",
"attributes": {
"email": "customer1@otenet.gr",
"# of policies": 4,
"policy_details": {
"POLICY_NO": "000000003",
"RECEIPT_NO": "420000003",
"PAYMENT_CODE": "RF35000000000000000000003",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000001",
"attributes": {
"email": "customer1@otenet.gr",
"# of policies": 4,
"policy_details": {
"POLICY_NO": "000000004",
"RECEIPT_NO": "420000004",
"PAYMENT_CODE": "RF35000000000000000000004",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000005",
"RECEIPT_NO": "420000005",
"PAYMENT_CODE": "RF35000000000000000000005",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000006",
"RECEIPT_NO": "420000006",
"PAYMENT_CODE": "RF35000000000000000000006",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000007",
"RECEIPT_NO": "420000007",
"PAYMENT_CODE": "null",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000008",
"RECEIPT_NO": "420000008",
"PAYMENT_CODE": "RF35000000000000000000008",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000009",
"RECEIPT_NO": "420000009",
"PAYMENT_CODE": "RF35000000000000000000009",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000002",
"attributes": {
"email": "customer2@gmail.com",
"# of policies": 6,
"policy_details": {
"POLICY_NO": "000000010",
"RECEIPT_NO": "420000010",
"PAYMENT_CODE": "null",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000003",
"attributes": {
"email": "customer3@yahoo.com.au",
"# of policies": 5,
"policy_details": {
"POLICY_NO": "000000011",
"RECEIPT_NO": "420000011",
"PAYMENT_CODE": "RF35000000000000000000011",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000003",
"attributes": {
"email": "customer3@yahoo.com.au",
"# of policies": 5,
"policy_details": {
"POLICY_NO": "000000012",
"RECEIPT_NO": "420000012",
"PAYMENT_CODE": "RF35000000000000000000012",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000003",
"attributes": {
"email": "customer3@yahoo.com.au",
"# of policies": 5,
"policy_details": {
"POLICY_NO": "000000013",
"RECEIPT_NO": "420000013",
"PAYMENT_CODE": "null",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000003",
"attributes": {
"email": "customer3@yahoo.com.au",
"# of policies": 5,
"policy_details": {
"POLICY_NO": "000000014",
"RECEIPT_NO": "420000014",
"PAYMENT_CODE": "RF35000000000000000000014",
"KLADOS": "Αυτοκινήτου"
}
}
}, {
"type": "customer",
"customer_id": "1-0000003",
"attributes": {
"email": "customer3@yahoo.com.au",
"# of policies": 5,
"policy_details": {
"POLICY_NO": "000000015",
"RECEIPT_NO": "420000015",
"PAYMENT_CODE": "RF35000000000000000000015",
"KLADOS": "Αυτοκινήτου"
}
}
}
]
< /code>
< /li>
< /ol>
< /li>
2nd Method < /p>
# Create the new 'attributes' field as dictionary
df['attributes'] = df[['email', '# of policies']].to_dict(orient='records')

# Convert the dictionary 'attributes' to a string
df['attributes'] = df[['attributes']].astype(str)

# JSON Section

# Create the JSON structure
j = (df.groupby(['type', 'customer_id', 'attributes'])
.apply(lambda x: x[['POLICY_NO', 'RECEIPT_NO','PAYMENT_CODE', 'KLADOS']].to_dict('records'), include_groups=False)
.reset_index()
.rename(columns={0:'policies details'})
.to_dict(orient='records')
)

# Convert list of dictionaries to JSON
json_output = json.dumps(j, indent=4, ensure_ascii=False)

print(json_output)
< /code>
На этот раз: < /p>

'type', 'customer_id' и «атрибуты» наконец -то сгруппированы, но «атрибуты» представлены как строка, а не как словарь (ключ: пары ценностей) < /p>
< /li>
plisters ', но

Подробнее здесь: https://stackoverflow.com/questions/795 ... in-a-repea