How to convert the pandas row to custom json format and make a POST request

Question

My DataFrame contains the following columns and rows

hosp  doc   dep  p_t    ins   tpa       p_n  i_c  t_date       cat    c_amt

ALZ   Dr.M  Onc  SAICO  SAISA AZBRONZE  AZS  11   2020-08-11   Cons   341.25
ALZ   Dr.K  Card Mitra  Mit   ASGOLD    ASG  8265 2020-08-15   Cons   1123.45

I want to convert each row to the following json format and make a post request to an API. (Please note that the "id" will always be 1)

{
    "hosp": "ALZ",
    "doc": "Dr.M",
    "dep": "Onc",
    "p_t": "SAICO",
    "ins": "SAISA",
    "tpa": "AZBRONZE",
    "p_n": "AZS",
    "activities": [
        {
            "id": "1",
            "i_c": "11",
            "t_d": "2020-08-11",
            "cat": "Cons",
            "c_amt": "341.25"
        }
    ]
}

Finally add the response as a new column to the DataFrame

hosp  doc   dep  p_t    ins   tpa       p_n  i_c  t_date       cat    c_amt    response

ALZ   Dr.M  Onc  SAICO  SAISA AZBRONZE  AZS  11   2020-08-11   Cons   341.25   Yes
ALZ   Dr.K  Card Mitra  Mit   ASGOLD    ASG  8265 2020-08-15   Cons   1123.45  No

Answer 1

You can apply pandas function https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_json.html with df.to_json() and then access each row to get the json presentation.

Answer 2

Here is solution

Since you would want to run this for all the rows present in the DataFrame, do the following

# Number of rows all_rows = len(df) for i in range(all_rows): # choose the first half of cols before "activities" x = dict(df.iloc[i, 0:7]) # add the "activites" with the rest of the cols x["activities"] = [dict(df.iloc[i, 7:])] # dict has single quotes which throws an error if sent to an API, so use json.dumps x = json.dumps(x)

To send a POST request to an API

url = 'http://your_api/api/v1/response' headers = { 'content-type': "application/json" } response = requests.request("POST", url, data=x, headers=headers) res = json.loads(response.content)

Answer 3

You can use this implementation,

1. Take the columns which you want to put under activities
2. Drop these columns from original dataframe
3. Set Id as you want
4. Convert the activities dataframe into json
5. Set activities json as activities column to original dataframe

columns = ['i_c', 't_date', 'cat', 'c_amt']
activities = df[columns]
df.drop(columns=columns, inplace=True)
activities['id'] = '1'
activities_json = eval(activities.to_json(orient='records'))
df['activities'] = activities_json