Question

我有一个以下格式的示例表，其中包含示例数据：

= cust_table

Customer:  Liked_Color:
Adam       Blue    
Adam       Green    
Adam       Yellow    
Adam       Red    
Bob        Yellow    
Bob        Yellow    
Bob        Yellow    
Bob        Yellow    
Charlie    Red    
Charlie    Red    
Charlie    Red    
Charlie    Red

我该如何选择不同的客户，并且仅在不喜欢蓝色的情况下才返回他们？

因此，返回的结果将是：

客户：Bob，Charlie

如果我这样做：

SELECT DISTINCT Customer
FROM cust_table
WHERE Liked_Color NOT LIKE Blue

我明白了：亚当，鲍勃，查理

我如何确保仅当Blue不是客户的Liked_Color时才返回客户？

Answer 1

我将使用聚合：

select customer
from cust_table
group by customer
having sum(case when color = 'Blue' then 1 else 0 end) = 0;

但是，customer表中每个客户应该有一行（至少每单位时间）。如果您有这样一个表和一个customer_colors表，那么我将使用not exists：

select c.*
from customers c
where not exists (select 1
                  from customer_colors cc
                  where cc.customer = c.customer and cc.color = 'Blue'
                 );

这也将使返回的客户根本没有首选的颜色，而这是您无法用单个表完成的。

Answer 2

或者，查看集合是否有好处：

SQL> with cust (customer, liked_color) as
  2    (select 'Adam'   , 'Blue'   from dual union all
  3     select 'Adam'   , 'Green'  from dual union all
  4     select 'Bob'    , 'Yellow' from dual union all
  5     select 'Bob'    , 'Red'    from dual union all
  6     select 'Charlie', 'Red'    from dual union all
  7     select 'Charlie', 'Red'    from dual
  8    )
  9  select customer from cust where liked_color <> 'Blue'
 10  minus
 11  select customer from cust where liked_color = 'Blue';

CUSTOME
-------
Bob
Charlie

SQL>

Answer 3

这将是所有方法中最简单的方法：

select distinct a from Table1 where a not in 
(select distinct a from Table1 where b like 'Blue');

检查小提琴：http://sqlfiddle.com/#!9/7034f/10

如何返回满足条件的不同行

3 个答案: