執行SQL查詢以排序列表，而不必將其存儲為Java對象？

Question

我在MySQL中有一個包含3列的表：CategoryID，Score，dateTime。 我需要找到得分最高的前4個類別。 我還必須考慮每個分數的時間戳，因為如果我在較舊的日期在某個類別中得分較高，而在最近的某個日期中該類別得分較低，則最近的分數將主導該類別的分數。 SQL中是否存在查詢以生成結果集？ 我在Java中通過將數據存儲為對象來完成此操作，但是如果有一種更有效的方法，我更願意這樣做。

注意：pickRandom是字符串的ArrayList，並且忽略方法checkIfThere（）

Scores prevScore = new Scores(0, "AD",  new Timestamp(0));
            while (rs.next()) {

                categoryID = rs.getString("categoryID");
                int relScore = rs.getInt("relevanceScore");
                Timestamp ts = rs.getTimestamp("date_time");

                Scores currScore = new Scores(relScore, categoryID, ts);

                // ensures that a Random Barrier Cateogry is prompted
                if ((currScore.getScore() >= prevScore.getScore() && (currScore.getDate().compareTo(prevScore.getDate()) > 0) && !checkIfThere(user, categoryID))) {
                    pickRandom.add(categoryID);
                    prevScore = currScore;
                }
            }

編輯：

表： 分數

'AD', '110', '2015-08-26 14:59:51' 
'DC', '110', '2015-08-26 14:59:51' 
'DO', '110', '2015-08-26 14:59:51' 
'EC', '8.087601288990395', '2015-08-26 14:59:53' 
'EC', '110', '2015-08-26 14:59:51'

Answer 1

嘗試以下操作。 inr派生表依靠唯一的時間/秒時間分段。 也許可以使用id列進行增強。 inr將在我的樣本數據中獲得所有6個類別。 外部將加入並說只需要4。

create table scores
(   id int auto_increment primary key,
    categoryID varchar(10) not null,
    relevanceScore float not null,
    date_time datetime not null
);
truncate table scores;
insert scores (categoryID,relevanceScore,date_time) values
('AD', '110', '2015-08-26 14:59:51'), 
('DC', '110', '2015-08-26 14:59:51'), 
('DO', '110', '2015-08-26 14:59:51'), 
('EC', '8.087601288990395', '2015-08-26 14:59:53'), 
('EC', '110', '2015-08-26 14:59:51'),
('OLD1', '110', '2014-08-26 14:59:51'),
('OLD2', '110', '2014-08-26 14:59:51');

選項A ：（按date_time DESC排在前4位）

select s.* from scores s 
join 
( select categoryID,max(date_time) as maxdt 
  from scores 
  group by categoryId 
) inr 
on inr.categoryID=s.categoryID and inr.maxdt=s.date_time 
order by s.date_time desc 
limit 4; 

+----+------------+----------------+---------------------+
| id | categoryID | relevanceScore | date_time           |
+----+------------+----------------+---------------------+
|  4 | EC         |         8.0876 | 2015-08-26 14:59:53 |
|  1 | AD         |            110 | 2015-08-26 14:59:51 |
|  2 | DC         |            110 | 2015-08-26 14:59:51 |
|  3 | DO         |            110 | 2015-08-26 14:59:51 |
+----+------------+----------------+---------------------+
4 rows in set (0.00 sec)

選項B：（relevanceScore DESC排名前4）

select s.* from scores s 
join 
( select categoryID,max(date_time) as maxdt 
  from scores 
  group by categoryId 
) inr 
on inr.categoryID=s.categoryID and inr.maxdt=s.date_time 
order by s.relevanceScore desc
limit 4; 
+----+------------+----------------+---------------------+
| id | categoryID | relevanceScore | date_time           |
+----+------------+----------------+---------------------+
|  1 | AD         |            110 | 2015-08-26 14:59:51 |
|  2 | DC         |            110 | 2015-08-26 14:59:51 |
|  3 | DO         |            110 | 2015-08-26 14:59:51 |
|  6 | OLD1       |            110 | 2014-08-26 14:59:51 |
+----+------------+----------------+---------------------+
4 rows in set (0.00 sec)

Answer 2

嘗試這個

---

select * from Scores order by date_time,relevanceScore;