Question

我有一张表，我需要为每个max(bar)选择foo 如果 <{strong> foo和bar低于另一条记录的foo和bar，则和会排除任何记录。< / p>

数据：

+------+------+ | foo | bar | +------+------+ | 5 | 30 | | 5 | 40 | | 6 | 60 | | 6 | 65 | | 6 | 95 | | 7 | 10 | | 7 | 30 | +------+------+

预期结果：

+------+----------+ | foo | max(bar) | +------+----------+ | 6 | 95 | | 7 | 30 | +------+----------+

我坚持

+------+----------+ | foo | max(bar) | +------+----------+ | 5 | 40 | | 6 | 95 | | 7 | 30 | +------+----------+

我尝试了以下查询的不同变体，但没有运气。

select foo, max(bar) from mytable group by foo order by foo, bar;

Answer 1

您的查询几乎就在那里。您只需要一个合适的过滤条件。以下是使用not exists进行过滤的一种方法：

select foo, max(bar)
from mytable t
where not exists (select 1
                  from mytable t2
                  where t.foo < t2.foo and t.bar < t2.bar
                 )
group by foo
order by foo;

Answer 2

带有LEFT JOIN的版本：

SELECT MAX(mt1.bar), mt1.foo 
FROM myTable mt1 
LEFT JOIN myTable mt2 ON (mt1.foo < mt2.foo AND mt1.bar < mt2.bar)
WHERE mt2.foo IS NULL
GROUP BY mt1.foo
;

SQL Fiddle

Answer 3

尝试获取子查询中每列的最大值，然后按

尝试您的组

select t1.foo, max(t1.bar) 
from mytable t1, (select max(foo) foo, max(bar) bar from mytable) t2 
where t1.foo = t2.foo 
or t1.bar = t2.bar 
group by foo;

SQL - 选择最多两列

3 个答案: