我有一个查询:
SELECT
CST_CON_ID as ID_USER,
X_TYPE_FOURTH_LVL as ELEMENT_TYPE,
COUNT(X_TYPE_FOURTH_LVL) as T_REPEAT
FROM
SIEBEL.S_SRV_REQ
GROUP BY
CST_CON_ID,
X_TYPE_FOURTH_LVL
ORDER BY
CST_CON_ID;
运行时会返回以下信息:
ID_USER ELEMENT_TYPE T_REPEAT
1-1HQR-1380 REPOSITION 3
1-1HQR-1380 CANCELED 4
1-1HQR-1380 INFORMATION 1
1-1HQR-1380 SELLS 90
1-1HQR-1380 PROMOTIONS 3
1-1HQR-1380 ACCOUNTS 5
1-1HQR-1250 ACTIONS 34
1-1HQR-1250 PROG 60
1-1HQR-1250 INFORMATION 3
1-1HQR-1560 SELLS 40
1-1HQR-1560 BUYS 2
1-1HQR-1560 REPOSITION 1
1-1HQR-1560 PROMOTIONS 70
但是现在我只想为每个ID_USER
提取T_REPEAT
的最大值。
我想要的预期输出将如下所示:
ID_USER ELEMENT_TYPE T_REPEAT
1-1HQR-1380 SELLS 90
1-1HQR-1250 PROG 60
1-1HQR-1560 PROMOTIONS 70
是否可以仅使用单个查询获得此输出?
答案 0 :(得分:1)
一种选择是使用row_number()
:
SELECT ID_USER, ELEMENT_TYPE, T_REPEAT
FROM (SELECT CST_CON_ID as ID_USER, X_TYPE_FOURTH_LVL as ELEMENT_TYPE,
COUNT(X_TYPE_FOURTH_LVL) as T_REPEAT,
ROW_NUMBER() OVER (PARTITION BY CST_CON_ID ORDER BY COUNT(X_TYPE_FOURTH_LVL) DESC) AS Seq
FROM SIEBEL.S_SRV_REQ
GROUP BY CST_CON_ID, X_TYPE_FOURTH_LVL
) t
WHERE Seq = 1;
编辑:如果您与count有联系,则使用dense_rank()
函数:
答案 1 :(得分:-1)
将当前语句包装在子查询中。
SELECT ID_USER, ELEMENT_TYPE, MAX(T_REPEAT)
FROM (
SELECT
CST_CON_ID as ID_USER,
X_TYPE_FOURTH_LVL as ELEMENT_TYPE,
COUNT(X_TYPE_FOURTH_LVL) as T_REPEAT
FROM
SIEBEL.S_SRV_REQ
GROUP BY
CST_CON_ID,
X_TYPE_FOURTH_LVL
ORDER BY
T_REPEAT
)
GROUP BY ID_USER