获取不同值的总计数

Question

我有一个包含3个字段的品牌，型号和BenchMark分数的MySQL表。

有关数据如何在此表中的注意事项。

1.品牌可以有不同的型号。 2.不同的品牌可以有同样的型号 3.特定品牌型号组合可以有不同的Benchmark分数。

所以表中的数据将是这个样子

Brand ModelName Benchmark 

B1 M1   1000 
B1 M1   2000 
B1 M2   3000 
B1 M2   2000 
B2 M1   3000 
B2 M3   4000 

现在我需要的是找到的时间数量的特定基准分数报告。

在这里，我们应该只计算那些模型 - 品牌组合的最大基准分数，而不是所有这些基准分数。

结果应该是这样的

Score Times 
3000 2 
2000 1 
4000 1 

那么应该是MySQL的声明，如果我需要这个result..please帮助

Answer 1

SELECT * FROM 
(SELECT MX,COUNT(1) SC FROM 
(SELECT MAX(MaxScores.MaxScore) MX,ScoreCounts.ScoreCount FROM 
(
    SELECT MAX(BenchMark) MaxScore,Brand,ModelName 
    FROM benchmarks GROUP BY Brand,ModelName 
) MaxScores 
INNER JOIN 
(
    SELECT COUNT(1) ScoreCount,Brand,ModelName 
    FROM benchmarks GROUP BY Brand,ModelName 
) ScoreCounts 
USING (Brand,ModelName) 
GROUP BY Brand,ModelName 
) A GROUP BY MX) AA ORDER BY SC DESC,MX; 

此查询完全产生你在找什么，并将结果显示在你的问题有它的确切顺序。我用这个问题的样本数据进行了试验。这里是我得到的：

mysql> use test 
Database changed 
mysql> drop table if exists benchmarks; 
Query OK, 0 rows affected (0.05 sec) 

mysql> CREATE TABLE benchmarks (Brand CHAR(2),ModelName CHAR(2),BenchMark INT,key (Brand,ModelName)); 
Query OK, 0 rows affected (0.14 sec) 

mysql> INSERT INTO benchmarks VALUES 
    -> ('B1','M1',1000), 
    -> ('B1','M1',2000), 
    -> ('B1','M2',3000), 
    -> ('B1','M2',2000), 
    -> ('B2','M1',3000), 
    -> ('B2','M3',4000); 
Query OK, 6 rows affected (0.03 sec) 
Records: 6 Duplicates: 0 Warnings: 0 

mysql> SHOW CREATE TABLE benchmarks\G 
*************************** 1. row *************************** 
     Table: benchmarks 
Create Table: CREATE TABLE `benchmarks` (
    `Brand` char(2) DEFAULT NULL, 
    `ModelName` char(2) DEFAULT NULL, 
    `BenchMark` int(11) DEFAULT NULL, 
    KEY `Brand` (`Brand`,`ModelName`) 
) ENGINE=InnoDB DEFAULT CHARSET=latin1 
1 row in set (0.00 sec) 

mysql> SELECT * FROM benchmarks; 
+-------+-----------+-----------+ 
| Brand | ModelName | BenchMark | 
+-------+-----------+-----------+ 
| B1 | M1  |  1000 | 
| B1 | M1  |  2000 | 
| B1 | M2  |  3000 | 
| B1 | M2  |  2000 | 
| B2 | M1  |  3000 | 
| B2 | M3  |  4000 | 
+-------+-----------+-----------+ 
6 rows in set (0.01 sec) 

mysql> SELECT * FROM 
    -> (SELECT MX,COUNT(1) SC FROM 
    -> (SELECT MAX(MaxScores.MaxScore) MX,ScoreCounts.ScoreCount FROM 
    -> (
    ->  SELECT MAX(BenchMark) MaxScore,Brand,ModelName 
    ->  FROM benchmarks GROUP BY Brand,ModelName 
    ->) MaxScores 
    -> INNER JOIN 
    -> (
    ->  SELECT COUNT(1) ScoreCount,Brand,ModelName 
    ->  FROM benchmarks GROUP BY Brand,ModelName 
    ->) ScoreCounts 
    -> USING (Brand,ModelName) 
    -> GROUP BY Brand,ModelName 
    ->) A GROUP BY MX) AA ORDER BY SC DESC,MX; 
+------+----+ 
| MX | SC | 
+------+----+ 
| 3000 | 2 | 
| 2000 | 1 | 
| 4000 | 1 | 
+------+----+ 
3 rows in set (0.01 sec) 

试试看！

Answer 2

SELECT b.model, b.brand, bd.score, COUNT(*) 
FROM (
     SELECT model, brand, MAX(score) AS mscore 
     FROM benchmark 
     GROUP BY 
       model, brand 
     ) bd 
JOIN benchmark b 
ON  b.model = bd.model 
     AND b.brand = bd.brand 
     AND b.score = bd.mscore 
GROUP BY 
     b.model, b.brand 

上(model, brand, score)这个工作快速创建索引。

更新：

SELECT score, COUNT(*) AS cnt 
FROM benchmark 
GROUP BY 
     score 
ORDER BY 
     cnt DESC