df = pd.read_csv("sample.csv")
列印（df）
| ID | 回復 | 日期 |
|---|---|---|
| 234 | {"statusCode":"00","statusDescription":"Successful","mRecord":"202111105530685","tranxReference":"012021116029","re??cipient":"09131976","amount":"1500","代碼":"202505651505637179","network":"MVP","tranxDate":"14-11-2021 10:50 am"} | 2021-11-14 10:50:55 |
| 235 | {"statusCode":"00","statusDescription":"Successful","mRecord":"2021111496980","tranxReference":"01202111057048","re??cipient":"091598","amount":"1500"," confirmCode":"D211114.1050040","network":"MVP","tranxDate":"14-11-2021 10:50 am"} | 2021-11-14 10:50:56 |
我想在另一個資料框中拆分“回應”列上的所有記錄，并且所有鍵都將用于列標題
uj5u.com熱心網友回復：
這可以分兩步完成：首先將記錄response分解，然后df在洗掉原始response列后連接。您可能需要申請json.loads，因為您的 json 記錄似乎被讀取為字串：
import json
df_response = pd.json_normalize(df['response'].apply(json.loads))
df_out = pd.concat([df.drop('response', axis=1), df_response], axis=1)
print(df_out)
輸出：
ID Date statusCode statusDescription ... Code network tranxDate confirmCode
0 234 2021-11-14 10:50:55 00 Successful ... 202505651505637179 MVP 14-11-2021 10:50 am NaN
1 235 2021-11-14 10:50:56 00 Successful ... NaN MVP 14-11-2021 10:50 am D211114.1050040
轉載請註明出處，本文鏈接：https://www.uj5u.com/qukuanlian/414790.html
標籤：