Question

想象一下，我有一个包含{id，username，firstname，lastname}

的客户的数据库表

如果我想找到有多少个不同名字的实例我可以做：

select firstname,count(*) from Customers group by 2 order by 1;

   username | count(*)
   ===================
   bob      |   1
   jeff     |   2
   adam     |   5

如何编写相同的查询以仅返回多次出现的名字？即在上面的例子中只返回jeff和adam的行。

Answer 1

您需要having子句，如下所示：

select 
    firstname,
    count(*) 
from Customers 
group by firstname
having count(*) > 1
order by 1

Answer 2

我应该说，{p> group by 2 order by 1很可怕。如果支持，请使用正确的列名：这将大大提高可读性。

考虑到这一点，

select firstname, count(*) c 
from Customers 
group by firstname  
having count(*) > 1 -- Kudos to Shannon
order by c;

Answer 3

这就是HAVING子句的作用。我不确定这是否适用于informix，但请试一试：

select firstname, count(*) 
from Customers 
group by firstname
HAVING COUNT(*) > 1

SQL - WHERE AGGREGATE＆gt; 1

3 个答案: