Question

我有下表：

row_number itemID   score
1           45      99
2           45      47
1           46      98
2           46      20
1           47      98
2           47      20
…           …        …

以下查询的输出是什么：

SET @num := 0, @itemID_grouping := ''; 

SELECT row_number, 
       itemID, 
       score 
FROM   (SELECT itemID, 
               score, 
               @num := IF(@itemID_grouping = itemID, @num + 1, 1) AS row_number, 
               @itemID_grouping := itemID                         AS temp 
        FROM   my_table 
        ORDER  BY itemID, 
                  score DESC) AS x 
WHERE  x.row_number <= 2;

来自以下原始表：

my_table: 
itemID score
 46     99
 …      …

因此，对于每个itemID，我有一组两个值（两个最高分）。

我想要另一个包含＆＃39; y＆＃39;或者，如果对于每个集合，row_number=1的数字大于某个阈值且row_number=2的数字低于某个阈值。有什么想法吗？

Answer 1

好的，我找到了一个答案（可能不是很快，但它可以完成工作）：

set @num := 0, @itemID_grouping := '',@num_temp:=0,@flag1:=0,@flag2:=0;

select row_number,itemID, score,flag2
from 
(
   select itemID,score,
      @num := if(@itemID_grouping = itemID, @num + 1, 1) as row_number,
      @flag1:= if(@num=1 AND score>90,1,0) as flag1,
      @flag2:= if(@num=2 AND @flag_temp=1 AND score<70,1,0) as flag2, 
      @itemID_grouping := itemID as temp,
      @flag_temp:=@flag1 as temp_flag
  from my_table
  order by itemID,score desc
) as x 
where x.row_number <= 2;

应用于MySQL中的行子集的条件列

1 个答案: