Question

我希望通过＆＃39; location＆amp ;;来选择最畅销的销售人员。产品＆＃39;组合。第一批我和排序排名，然后我总结为＆＃34;位置＆amp;产品＆＃34;，然后我想放弃除了顶级销售人员之外的所有＆＃34;位置＆amp;产物＆＃34;这样查询结果只包含顶级销售人员的位置＆amp;位置＆amp;产物＆＃34;每个＆＃34;位置＆amp;的总销售额产物＆＃34;组合

我不具备的一部分是＆＃34;选择Rank = 1＆＃34; - 添加＆＃34;其中＆＃34; ＆＃34;来自＆＃34;不起作用：

Select Location, Product, Sales, Salesperson
       row_number() OVER (PARTITION BY concat(Location,Product) ORDER BY Sales desc) AS Rank,
       Sum(Sales) OVER(PARTITION BY concat(Location,Product)) AS Salesbylocprod
From Table A

Answer 1

您可以使用子查询。此外，您不需要连接partition by：

的密钥

Select Location, Product, Sales, Salesperson, Salesbylocprod
from (select a.*
             row_number() over (partition by Location, Product order by Sales desc) as seqnum,
             sum(Sales) over (partition by Location, Product) as Salesbylocprod
      From Table A
     ) a
where seqnum = 1;

请注意，我已将列的名称从rank更改为seqnum。 rank有点暗示它是使用rank()函数计算的。这与row_number()不同，因此该名称具有误导性。

Answer 2

您还可以使用公用表格表达式：

;WITH CTE 
     AS (SELECT *,
                ROW_NUMBER() OVER(PARTITION BY Location,Product ORDER BY Sales DESC) AS RowNum,
                SUM(Sales) OVER(PARTITION BY Location,Product) AS SalesByLocProd
         FROM TableA
        )
        SELECT Location,
             Product,
             Sales,
             SalesPerson,
             SalesByLocProd
        FROM CTE
        WHERE RowNum = 1;

在select语句中 - 如何在函数b4之后包含一个过滤器来自from子句

2 个答案: