Question

我有一个表“vote_benefits”，其中包含以下架构： id（int 11） id_benefit（int 11）评级（int 11）手机（varchar 10） inputdate（日期）

人们可以为每件物品投票（通过星级评分系统），从1到5星。我必须为每项福利获得每个评级的投票数。

内容类似于

id id_b rating cellphone  inputdate
1   2   5   123456789   2011-01-04 18:56:38
2   2   4   123456789   2011-01-04 19:03:27
3   2   4   123456789   2011-01-05 10:24:29
4   2   4   123456789   2011-01-05 10:24:33
5   2   5   123456789   2011-01-05 10:24:37
6   2   5   123456789   2011-01-05 10:24:42
7   2   4   123456789   2011-01-05 10:24:43
8   2   5   123456789   2011-01-05 10:24:44
9   2   4   123456789   2011-01-05 10:24:45
10  2   5   123456789   2011-01-05 10:24:48
11  2   5   123456789   2011-01-05 10:25:42
12  3   4   123456789   2011-01-05 10:49:20
13  3   5   123456789   2011-01-05 10:49:21
14  3   4   123456789   2011-01-05 10:49:22
15  3   5   123456789   2011-01-05 10:52:44
16  3   5   123456789   2011-01-05 10:52:45
17  3   4   123456789   2011-01-05 10:53:04
18  3   4   123456789   2011-01-05 10:53:05

我希望从这些按照id_benefit分组的最多投票权益中获益。我到目前为止

SELECT id_benefit, count( * ) AS votes, rating
FROM `vote_benefits 
GROUP BY id_benefit, rating

给了我

id_b votes rating
2   5      4
2   6      5
3   4      4
3   3      5

但是我想通过id_b对这些进行分组，获得每个id_b投票最多的行：

id_b votes rating
2      6     5
3      4     4

我读过像SQL - select distinct records in one field with highest records from another field这样的链接，但我不明白如何根据我的需要调整它。

Answer 1

我相信以下内容应该有效：

SELECT id_b, MAX(votes), rating 
FROM (SELECT id_benefit, count( * ) AS votes, rating
        FROM `vote_benefits 
    GROUP BY id_benefit, rating) 
GROUP BY id_b

Answer 2

select 
id_b, max(rating_count), rating 
from 
(
  select id_b, rating, count(*) as rating_count 
  from vote_benefits 
  group by id_b, rating 
) as tbl
group by id_b;

sql：从每个用户中选择最多的投票项目

2 个答案: