Pandas - 将CSV转换为JSON - 使用groupby进行组织，然后删除索引键

Question

I have a CSV file that I'm trying to convert to JSON with Pandas. It has multiple headers but for the sake of simplicity let's just say it has 3: "region", "state", and "salesperson". Three columns, three rows that often have repeating values within (state names and such).

My ideal result is:

{
    "salesperson": [
        {
            "name": "John Doe",
            "values": [
                {
                    "region": "North America",
                    "state": "Connecticut"
                },
                {
                    "region": "North America",
                    "state": "Vermont"
                }
            ]
        },
        {
            "name": "Jane Doe",
            "values": [
                {
                    "region": "North America",
                    "state": "New York"
                },
                {
                    "region": "North America",
                    "state": "New Hampshire"
                }
            ]
        }
    ]
}

This is what I currently have for reading the data and turning it into JSON.

df = pd.read_csv('Foo.csv', encoding="ISO-8859-1",
                    escapechar='\\')
result = (df.groupby(['salesperson'])
            .apply(lambda x: x.to_dict('r'))
            .to_json(orient='table')
            )
return result

.to_json(orient='table') is close, it gives me

"data": [
    {
        "salesperson": "John Doe",
        "values": [
            {
                "region": "North America",
                "state": "Connecticut",
                "salesperson": "John Doe"
            },

However "salesperson" is still in the "values". I've tried

result = (df.groupby(['salesperson'])
            .apply(lambda x: x.to_dict('r'))
            .drop('salesperson')
            .to_json(orient='table')
            )

But that doesn't seem to be the correct way.

I'm not sure how to tell it to use "salesperson" as the index & remove it from the output, without actually editing the JSON file after it's created.

Answer 1

The code below deletes the key not needed.

Step-1:

Assign a variable.

data = {
        "salesperson": "John Doe",
        "values": [
            {
                "region": "North America",
                "state": "Connecticut",
                "salesperson": "John Doe"
            }]
       }

Step-2: Delete key

del data['salesperson']

Output:

Answer 2

I had to apply drop to the key before applying to_dict()

result = df.groupby(df.salesperson).apply(
    lambda x: x.drop('salesperson', 1).to_dict('records')).to_json(orient='index')

This removed the key from the resulting JSON values while preserving it as the index.