HAVING 子句中的多个聚合函数

Posted 2023-02-24

【中文标题】HAVING 子句中的多个聚合函数

【英文标题】：Multiple aggregate functions in HAVING clause

【发布时间】：2013-01-23 06:05:40

【问题描述】：

由于我查询的性质，我有计数为 3 的记录，这些记录也符合计数为 2 的条件，依此类推。我想知道是否可以查询“计数大于 x 且小于 7”？我怎么会写这个。这是我当前的代码。

GROUP BY meetingID
HAVING COUNT( caseID )<4

我想要类似的东西

GROUP BY meetingID
HAVING COUNT( caseID )<4 AND >2

这样的话，它只计算 3

【参考方案1】：

GROUP BY meetingID
HAVING COUNT(caseID) < 4 AND COUNT(caseID) > 2

【参考方案2】：

对于您的示例查询，大于 2 且小于 4 的唯一可能值是 3，因此我们简化：

GROUP BY meetingID
HAVING COUNT(caseID) = 3

---

在您的一般情况下：

GROUP BY meetingID
HAVING COUNT(caseID) > x AND COUNT(caseID) < 7

或者（可能更容易阅读？），

GROUP BY meetingID
HAVING COUNT(caseID) BETWEEN x+1 AND 6

【参考方案3】：

不需要做两次检查，为什么不检查count = 3：

GROUP BY meetingID
HAVING COUNT(caseID) = 3

如果你想使用多重检查，那么你可以使用：

GROUP BY meetingID
HAVING COUNT(caseID) > 2
 AND COUNT(caseID) < 4

【参考方案4】：

这样的？

HAVING COUNT(caseID) > 2
AND COUNT(caseID) < 4

【参考方案5】：

select CUSTOMER_CODE,nvl(sum(decode(TRANSACTION_TYPE,'D',AMOUNT)),0)) DEBIT,nvl(sum(DECODE(TRANSACTION_TYPE,'C',AMOUNT)),0)) CREDIT,
nvl(sum(decode(TRANSACTION_TYPE,'D',AMOUNT)),0)) - nvl(sum(DECODE(TRANSACTION_TYPE,'C',AMOUNT)),0)) BALANCE from TRANSACTION   
GROUP BY CUSTOMER_CODE
having nvl(sum(decode(TRANSACTION_TYPE,'D',AMOUNT)),0)) > 0
AND (nvl(sum(decode(TRANSACTION_TYPE,'D',AMOUNT)),0)) - nvl(sum(DECODE(TRANSACTION_TYPE,'C',AMOUNT)),0))) > 0

【讨论】：

欢迎来到Stack Overflow！最好解释一下你做了什么。你能把它包括在你的答案中吗？谢谢！

【参考方案6】：

我正在写完整的查询，这将消除您的所有疑虑

SELECT BillingDate,
       COUNT(*) AS BillingQty,
       SUM(BillingTotal) AS BillingSum
FROM Billings
WHERE BillingDate BETWEEN '2002-05-01' AND '2002-05-31'
GROUP BY BillingDate
HAVING COUNT(*) > 1
AND SUM(BillingTotal) > 100
ORDER BY BillingDate DESC