select 重复行之一，其他行保持为 NULL SQL Netezza

Question

请在下面 table 中更改某些列的行，而其他一些列是重复的，因此我需要获取更改后的行并仅保留重复行中的一条记录和让其他人为 NULL

LOG_DATE  |ITEM_ID |PROBLEM |CALL_DATE  |UCID     |ID       |ACCOUNT_NO|CALL_STATUS|
----------|--------|--------|-----------|---------|---------|----------|-----------|
2020-10-23|76409075|TYPE 1  |2020-10-23 |603405210|133283340|3825551   |Answered   |
2020-10-07|76044107|TYPE 2  |2020-10-23 |603405210|133283340|3825551   |Answered   |
2020-10-18|76297468|TYPE 3  |2020-10-23 |603405210|133283340|3825551   |Answered   |

预期结果

LOG_DATE  |ITEM_ID |PROBLEM |CALL_DATE  |UCID     |ID       |ACCOUNT_NO|CALL_STATUS|
----------|--------|--------|-----------|---------|---------|----------|-----------|
2020-10-23|76409075|TYPE 1  |2020-10-23 |603405210|133283340|3825551   |Answered   |
2020-10-07|76044107|TYPE 2  |NULL       |NULL     |NULL     |NULL      |NULL       |
2020-10-18|76297468|TYPE 3  |NULL       |NULL     |NULL     |NULL      |NULL       |

Answer 1

我不认为我真的推荐这个，但你可以这样做：

select LOG_DATE, ITEM_ID, PROBLEM,
       (case when row_number() over (partition by CALL_DATE, UCID, ID, ACCOUNT_NO, CALL_STATUS order by log_date desc) = 1
             then call_date
        end) as call_date,
       (case when row_number() over (partition by CALL_DATE, UCID, ID, ACCOUNT_NO, CALL_STATUS order by log_date desc) = 1
             then ucid
        end) as ucid,
       . . .
from t;

select 重复行之一，其他行保持为 NULL SQL Netezza

select one of the duplicate rows and keep the others to be NULL SQL Netezza

sql

netezza