Question

我认为这对小组来说是普遍感兴趣的，因为这是一个实际的求职面试SQL问题，我错了。我想我很接近，但有人可以提供正确答案吗？

Brand  |  Model  |  Price
Braun  |  KF7150 |  98
Braun  |  KF7000 |  70
Braun  |  KF400  |  55
Krups  |  KM730  |  67
Krups  |  KM4689 |  130
Krups  |  EC311  |  50

对于上表（咖啡），请提供显示品牌，型号，价格以及每个型号价格与其品牌最高价格的比率的代码，四舍五入到最接近的2位小数。

我知道这段代码是错误的，因为它只选择整体最高价格（130）。

SELECT Brand, Model, Price, ROUND(Price/(SELECT MAX(Price) FROM Coffee),2)
 as Price_to_Brand_Highest
FROM Coffee
ORDER BY Brand;

我尝试了自我加入，但所有比率都回归为1，因为每个价格都是自己分开的。

SELECT C1.Brand, C1.Model, C1.Price, ROUND(C1.Price/(SELECT MAX(C2.Price) 
    FROM Coffee where C1.Brand=C2.Brand),2) as Price_to_Brand_Highest
FROM Coffee C1
JOIN Coffee C2 on C1.Model=C2.Model
GROUP BY C1.Model
ORDER BY C1.Brand;

Answer 1

您没有指定DBMS，因此这是ANSI标准SQL：

select brand, model, price, 
       round(price / max(price) over (partition by brand), 2) as price_ratio
from coffee
order by brand, model;

在线示例：http://rextester.com/HFZZRP41164

Answer 2

在将窗口函数添加到标准SQL之前，一种方法是连接一个计算最大值的子查询。

SELECT 
C1.*, 
ROUND(C1.Price/C2.MaxBrandPrice, 2) as Price_to_Brand_Highest
FROM Coffee C1
JOIN (
   Select Brand, max(Price) as MaxBrandPrice
   From Coffee
   Group by Brand
) C2
on (C1.Brand = C2.Brand)
ORDER BY C1.Brand, C1.Model;

Answer 3

; with Max_v as (
    Select Brand, max(Price) as MaxBrandPrice
        From Coffee
        Group by Brand
)
select Coffee.*, Max_v.price/(Max_ps.MaxBrandPrice*1.0)
    from Coffee, Max_v
    where D.Disease_ID = Max_ps.Disease_ID
    order by D.Disease_ID, D.ps desc

SQL的值与组的最大列值的比率

3 个答案: