JSON_Normalize Pandas 與列表

Question

我有一個按以下嵌套方式格式化的 JSON 文件。

[
  {
    "unitCode": "ABCD",
    "bedType": "Adult MT/MS",
    "census": 13,
    "subCensus": null,
    "censusDetails": [],
    "occupancy": 62,
    "occupancyStar": null,
    "occupancyAlertStatus": null,
    "columns": [
      {
        "id": "blockedBeds",
        "value": "1",
        "hoverDetails": [
          {
            "id": "bedName",
            "value": "23_1"
          }
        ]
      },
      {
        "id": "unOccupied",
        "value": "2",
        "hoverDetails": [
          {
            "id": "bedName",
            "value": "20a_2"
          },
          {
            "id": "bedName",
            "value": "22a_1"
          }
        ]
      }
    ],
    "codeEvents": null,
    "codeEventDetails": null
  },
  {
    "unitCode": "EFGH",
    "bedType": "Adult MT/MS",
    "census": 14,
    "subCensus": null,
    "censusDetails": [],
    "occupancy": 61,
    "occupancyStar": null,
    "occupancyAlertStatus": null,
    "columns": [
      {
        "id": "blockedBeds",
        "value": "1",
        "hoverDetails": [
          {
            "id": "bedName",
            "value": "52_2"
          }
        ]
      },
      {
        "id": "unOccupied",
        "value": "1",
        "hoverDetails": [
          {
            "id": "bedName",
            "value": "53_1"
          }
        ]
      }
    ],
    "codeEvents": null,
    "codeEventDetails": null
  }
]

我正在嘗試展平此文件並將其轉換為帶有 json_normalize 的json_normalize 。 這是我的代碼： testhover = json_normalize(data, ['columns'],['unitCode'])

我得到的 dataframe 如下所示：

    id          | value |   hoverDetails                                       | unitCode
0   blockedBeds | 1     |   [{'id': 'bedName', 'value': '23_1'}]               | ABCD
1   unOccupied  | 2     |   [{'id': 'bedName', 'value': '20a_2'}, {'id': '...' | ABCD
2   blockedBeds | 1     |   [{'id': 'bedName', 'value': '52_2'}]               | EFGH
3   unOccupied  | 1     |   [{'id': 'bedName', 'value': '53_1'}]               | EFGH

我需要以下格式：

    blockedBeds   |  unOccupied  |   unitCode
0 | '23_1'        |  NaN         |   ABCD
1 | NaN           |  '20a_2'     |   ABCD
2 | NaN           |  '22a_1'     |   ABCD
3 | '52_2'        |  NaN         |   EFGH
4 | NaN           |  '53_1'      |   EFGH

我似乎無法獲得嵌套床數據。 我非常感謝您的幫助。

Answer 1

您應該從循環中創建一個字典列表，並使用它來創建 dataframe。

vals = []

for item in parsed_json:
    unit_code = item['unitCode']
    for col in item['columns']:
        for hd in col['hoverDetails']:
            vals.append({'unitCode': unit_code,
                          col['id']: hd['value']})

pd.DataFrame(vals)

Output

  unitCode blockedBeds unOccupied
0     ABCD        23_1        NaN
1     ABCD         NaN      20a_2
2     ABCD         NaN      22a_1
3     EFGH        52_2        NaN
4     EFGH         NaN       53_1