JSON DataFrame 的包装，列中的每个唯一 ID

Question

The sample dataframe, df

SITE_ID     PRO_ID   PRO_ID     TXN_ID      LINE_ID         INST_ID     QUOTE       N1  N2  N3  R1  R2  R3
93672863K   PR_I     T_ID_PORT  283747E11   439329095       254553919   DISCOUNT    1   2   3   6   8   9
93672863K   PR_PI    T_PIP_COS  283747E12   8123619000      200613005   DISCOUNT    2   7   3   3   6   7
93672863K   PR_PI    T_PIP_PORT 283747E13   8123618999      200613003   DISCOUNT    6   5   9   1   5   9
93672863K   PR_PI    T_PIP_PORT 283747E14   8123618999      200613003   DISCOUNT    3   5   7   7   5   3
93672863K   PR_I     T_ID_PORT  283747E11   439329095       254553919   N-DISCOUNT  1   2   3   4   8   6
93672863K   PR_PI    T_PIP_COS  283747E12   8123619000      200613005   N-DISCOUNT  2   7   3   1   5   3
93672863K   PR_PI    T_PIP_PORT 283747E13   8123618999      200613003   N-DISCOUNT  6   5   9   8   4   2
93672863K   PR_PI    T_PIP_PORT 283747E14   8123618999      200613003   N-DISCOUNT  3   5   7   6   8   4

I'm trying to pack the dataframe to a JSON file using the below code

JSON_Dict = {"siteID": df.SITE_ID[0],
              "status": 1,
              "Message": None}
    
detail_LIST = []
        
for i in range(0, df.nunique()):
     detail_Dict_i = {"instID": df.INST_ID[i],
                       "ItemID": df.LINE_ID[i],
                       "opticount": [
                                      #N
                                     {"N1": df.N1[i],
                                      "N2": df.N2[i]
                                      "N3": df.N3[i]}

                                      #R
                                     {"R1": df.R1[i],
                                      "R2": df.R2[i]
                                      "R3": df.R3[i]}

                                    ]
                      }
                                
        
    detail_LIST.append(detail_Dict_i)

JSON_Dict["InstDetail"] = detail_LIST

This worked for me for the first four rows and without QUOTE column.

When the Quote column was added, the R1,R2,R3 changes with the Quote type. Now I'm trying include the R-Type columns with the both QUOTE types. As shown in the below

{
"siteID": 93672863K,
 "status": 1,
 "Message": None,

"InstDetail" :[
                {"instID": 254553919,
                  "ItemID": 439329095,
                  "opticount": [
                                 #N
                                {"N1": 1,
                                 "N2": 2,
                                 "N3": 3},

                                 #R with DISCOUNT
                                {"Quote": "DISCOUNT",
                                "R1": 6,
                                 "R2": 8,
                                 "R3": 9},

                                 #R with N -DISCOUNT
                                {"Quote": "N-DISCOUNT",
                                "R1": 4,
                                 "R2": 8,
                                 "R3": 6}

                                ]
                 },

                 {"instID": 200613005,
                  "ItemID": 8123619000,
                  "opticount": [
                                 #N
                                {"N1": 2,
                                 "N2": 7,
                                 "N3": 3},

                                 #R with DISCOUNT
                                {"Quote": "DISCOUNT",
                                "R1": 3,
                                 "R2": 6,
                                 "R3": 7},

                                 #R with N -DISCOUNT
                                {"Quote": "N-DISCOUNT",
                                "R1": 1,
                                 "R2": 5,
                                 "R3": 3}

                                ]
                 }, # other records


]


 }

I cound't find the logic to pack the JSON with each INST_ID with N columns and R columns with the QUOTE type.

I'm open to new ideas and approaches.

Answer 1

You can do it easily using groupby() :

result = {
    'siteID': df.SITE_ID[0],
    'status': 1,
    'Message': None,
    'InstDetail': [],
}

for items, group_df in df.groupby(['INST_ID', 'LINE_ID', 'N1', 'N2', 'N3']):
    inst_id, line_id, n1, n2, n3 = items
    detail = {
        'instID': inst_id,
        'ItemID': line_id,
        'opticount': [{
            'N1': n1,
            'N2': n2,
            'N3': n3,
        }]
    }

    for rec in group_df.to_dict('records'):
        detail['opticount'].append({
            'Quote': rec['QUOTE'],
            'R1': rec['R1'],
            'R2': rec['R2'],
            'R3': rec['R3'],
        })

    result['InstDetail'].append(detail)

print(result)