Question

我有一个看似微不足道的问题，但我无法使其发挥作用。它在Oracle SQL中。以下是脚本示例：

create table product (
product_id number primary key,
name varchar(255)
);

create table producer (
producer_id number primary key,
name varchar(255)
);

create table catalog (
pp_product_id number,
pp_producer_id number,
price number
);


alter table catalog add constraint pp_product_id1 foreign key (pp_product_id) references product (product_id);
alter table catalog add constraint pp_product_id2 foreign key (pp_producer_id) references producer (producer_id);


insert into product (product_id, name) values (1, 'HDD 250 gb');
insert into product (product_id, name) values (2, 'HDD 500 gb');
insert into product (product_id, name) values (3, 'HDD 750 gb');

insert into producer (producer_id, name) values (1, 'Hitachi');
insert into producer (producer_id, name) values (2, 'Corsair');
insert into producer (producer_id, name) values (3, 'Western Digital');

insert into catalog (pp_product_id, pp_producer_id, price) values (1,1, 80);
insert into catalog (pp_product_id, pp_producer_id, price) values (1,3, 60);
insert into catalog (pp_product_id, pp_producer_id, price) values (2,1, 75);
insert into catalog (pp_product_id, pp_producer_id, price) values (2,2, 40);
insert into catalog (pp_product_id, pp_producer_id, price) values (3,2, 63);
insert into catalog (pp_product_id, pp_producer_id, price) values (3,3, 100);

因此，共有6种产品，3种生产商，3种产品（HD类型）。我需要的是显示产品的CHEAPEST，每种产品类型与该产品相关的生产者名称。类似的东西：

Western Digital 250 gb 60

Corsair 500 gb 40

Corsair 750 gb 63

这将在他们的组中选择最便宜的HD（类型）

select p.name, min (c.price)
from product p, catalog c, producer prc
where c.pp_product_id = p.product_id and prc.producer_id = c.pp_producer_id
group by p.name;

...但我不能添加生产者名称？我尝试了几种解决方案，但都没有。

Answer 1

这应该回答你的核心问题。该帖子解决了MySQL中的问题，但适用相同的概念。

http://www.xaprb.com/blog/2006/12/07/how-to-select-the-firstleastmax-row-per-group-in-sql/

一旦你获得该组中的最低价格产品，剩下的就是加入。

Answer 2

这对row_number()很有用。您的数据没有type字段，因此我在产品表中发明了一个：

select *
from (select p.name as product_name, p.type, prod.name as producer_name,
             price,
             ROW_NUMBER() over (partition by p.name order by price ) as seqnum
      from catalog c join
           product p
           on c.pp_product_id = p.product_id join
           producer prod
           on c.pp_producer_id = prod.producer_id
     ) t
where seqnum = 1;

如果您想要所有产品的最低价格，请使用以下轻微变化：

select *
from (select p.name as product_name, p.type, prod.name as producer_name,
             price,
             min(price) over (partition by p.name) as minprice
      from catalog c join
           product p
           on c.pp_product_id = p.product_id join
           producer prod
           on c.pp_producer_id = prod.producer_id
     ) t
where price = minprice;

Answer 3

感谢@Gordon Linoff - 他的代码略有改动，它适用于我的问题。为了清楚起见，我在这里发布解决方案：

select *
from (select p.name as product_name, prod.name as producer_name,price,
ROW_NUMBER() over (partition by p.name order by price ) as seqnum
from catalog c join
product p
on c.pp_product_id = p.product_id join
producer prod
on c.pp_producer_id = prod.producer_id
) t
where seqnum = 1;

Answer 4

与此同时，我找到了另一个可能的答案：

SELECT p.name, pr.name, c2.minprice
FROM product p, producer pr, catalog c,
  (SELECT pp_product_id, MIN(price) AS minprice
    FROM catalog c1
    GROUP BY pp_product_id
  ) c2
WHERE p.product_id=c.pp_product_id
AND pr.producer_id=c.pp_producer_id
AND p.product_id  =c2.pp_product_id
AND c.price       =c2.minprice;

输出如下：

HDD 750 gb Corsair 6300

HDD 500 gb Corsair 4000

HDD 250 gb Western Digital 6000

如何从类别和生产者组中选择具有最低价格的产品？

4 个答案: