Pandas Dataframe row as a formatted JSON output - json

Question

I have data frame which I am trying to group by customer and print an output ,the to_json is not giving the format. Also I need to create separate json file for each customer, I think using the pandas generic method custom json formatting is not possible, what should be the direction I should be looking for.
I tried to group by customer_id , first_name and last_name and then set them as index and tried the orientation as index value but that didn't really worked out.
import pandas as pd
data = [{'customer_id': 1, 'first_name':'John', 'last_name':'Doe', 'amount':100, 'sub_amount':50,'total': 150,'product':'tool box'},
{'customer_id': 1, 'first_name':'John', 'last_name':'Doe', 'amount':50, 'sub_amount':50,'total': 100,'product':'light'},
{'customer_id': 2, 'first_name':'Jane', 'last_name':'Doe', 'amount':200, 'sub_amount':50,'total': 250,'product':'iron box'},
{'customer_id': 2, 'first_name':'Jane', 'last_name':'Doe', 'amount':50, 'sub_amount':50,'total': 100,'product':'led'}
]
df = pd.DataFrame(data)
df
customer_id first_name last_name amount sub_amount total product
0 1 John Doe 100 50 150 tool box
1 1 John Doe 50 50 100 light
2 2 Jane Doe 200 50 250 iron box
3 2 Jane Doe 50 50 100 led
expected output
{
"frist_name": "John",
"last_name": "Doe",
"Product_Details": {
"too box": {
"total": 150,
"amount": 100
},
"light": {
"total": 100,
"amount": 50
}
}
}

Answer 1

clients={}
for index,row in df.iterrows():
clients.setdefault(row['customer_id'], {'first_name': row['first_name'],
'last_name': row['last_name']})
clients[row['customer_id']].setdefault('Product_Details',{})[row['product']] = \
{'total': row['total'], 'amount': row['amount']}
print(json.dumps(clients[1],indent=4))