Question

表1：关键字

ItemID  KWtype   KWvalue

1       Brand     Nike
2       Brand     Nike
1       Type      Soccer
4       Brand     adidas

表2：项目

ItemID  SiteID  CategoryID  ItemName

1        0         10         Shoe1
2        0         10         Shoe2
3        2         55         Shoe3
4        1         11         Shoe4

我有这两张桌子（他们当然要大得多）

如何继续了解Site = 0和Category = 10中最常出现的10个品牌？我想在一列中列出前10名（排名前10位的品牌），在另一栏中，我想计算它们出现的次数

由于

Answer 1

首先，您应该真正调整数据模型。它可以更有效地工作，更容易理解

表：品牌

ID  ItemID  BrandName

 1       1  Nike
 2       2  Nike
 3       4  adidas

表：类型

ID  ItemID  TypeName

 1       1  Soccer

表：项目

ID  SiteID  CategoryID  ItemName

 1       0          10  Shoe1
 2       0          10  Shoe2
 3       2          55  Shoe3
 4       1          11  Shoe4

构建外键，将品牌和类型表中的ItemID链接到items表中的ID。

然后你可以运行像这样的查询

SELECT brandname as Brand, count(brand.id) as Qty
FROM brands
LEFT JOIN items ON items.id = brand.itemid
WHERE siteID = 0 AND categoryID = 10
GROUP BY brandname
ORDER BY qty DESC
LIMIT 10

Answer 2

这将计算：

SELECT k.KWvalue as Brand, count(*) as Qty
 FROM items as i
 JOIN keywords as k
   ON i.itemid = k.itemid
WHERE k. KWtype = Brand
  AND i.siteID = 0
  AND i.categoryID = 10
GROUP BY k.KWvalue

返回前10个计数取决于您的DBMS。

Darren H已经为mysql编写了它，只需添加

ORDER BY qty DESC
LIMIT 10

Teradata：

SELECT k.KWvalue as Brand, count(*) as Qty
 FROM items as i
 JOIN keywords as k
   ON i.itemid = k.itemid
WHERE k. KWtype = Brand
  AND i.siteID = 0
  AND i.categoryID = 10
GROUP BY k.KWvalue
QUALIFY RANK() OVER (ORDER BY Qty DESC) <= 10

Oracle（以及大多数其他DBMS）：

SELECT Brand, Qty
FROM
 (
   SELECT k.KWvalue AS Brand, COUNT(*) AS Qty, RANK() OVER (ORDER BY Qty DESC) AS rnk
    FROM items AS i
    JOIN keywords AS k
      ON i.itemid = k.itemid
   WHERE k. KWtype = Brand
     AND i.siteID = 0
     AND i.categoryID = 10
   GROUP BY k.KWvalue
 ) dt 
WHERE rnk <= 10

连接两个表进行计数，列表和排序

2 个答案: