Question

我有一张随机生成的表格。它有3列：P_Id，日期（插入行的位置）和文本。让我们说文本可以是各种动物。所以表格看起来像

1      2011-06-15 03:04:05    cow  
2      2011-06-15 03:04:15    dog  
3      2011-06-15 03:04:19    bird  
4      2011-06-15 03:04:30    cow  
5      2011-06-15 03:04:30    dog  
6      2011-06-15 03:04:35    cow  
....

挑选所有在最近进入的最后30分钟内至少发生过两次的动物的查询是什么？例如，这里的牛在最近一次进入的最后30分钟内发生了3次，即3:04:35，并且发生了两次。不会选择Bird，因为它只发生一次。

Answer 1

总是有直接的方式（即使用子查询来提取时间戳范围然后照常进行）：

select name, count(name)
from your_table
where created_at >= (
    select max(created_at) - INTERVAL 30 MINUTE
    from your_table
)
group by name
having count(name) >= 2

Answer 2

假设你检查每个事件的最后30分钟

select name, count(name)
from your_table t1
join (
    select name, max(created_at) created_at
    from your_table t2
) on t1.name = t2.name
where datediff (minute,t1.created_at,t2.created_at)<=30
group by name
having count(name) >= 2

Answer 3

select animal, count(*) 
from MyTable 
where date >= (currentDate - 30 minutes) 
group by animal
having count(*) >= 2

Answer 4

形成一个名为latest_animals的内部表，它将计算每只动物的最新进入日期。形成另一个名为previous_animals的内部表，它将计算每只动物的上一个最新进入日期。

根据动物名称在这两个表之间进行左联接，并根据两个日期之间的时差来过滤记录。

SELECT latest_animals.animal_name, TIME_DIFF(latest_animals.created_at, previous_animals.created_at) created_diff FROM

(
  SELECT animal_name, MAX(date) created_at FROM animals GROUP BY animal_name
) latest_animals
LEFT JOIN
(
  SELECT animal_name, MAX(date) created_at FROM animals GROUP BY animal_name WHERE id NOT IN
  (SELECT id, MAX(date) FROM animals GROUP BY animal_name)
) previous_animals
ON latest_animals.animal_name =  previous_animals.animal_name
WHERE created_diff > INTERVAL 30 MINUTE

需要帮助mysql查询

4 个答案: