SQL - 在案例中选择

Question

我有一个代码可以为“移动”的宇宙产生所需的输出，在这个宇宙中它查看每个周期并给我信息。我想要另一个产生相同输出但选择不同的查询，a “当前”宇宙。

当前的 Universe 会像现在一样计算总和，但仅针对在数据库中最晚日期符合标准的公司。

目前，该代码为 Market_Cap 在 0 到 10000 2015 之间的公司生成 EBIT/Sales 2015，依此类推，但我更希望它为当前 Market_Cap 在 0 到 10000 之间的公司生成 EBIT/Sales 2015（最新 date_month在 Market_Cap 中）

我正在使用 Microsoft SQL Server Management Studio

我尝试在 sum(case when ... then) 语法中插入另一个标准，如“And c.company_id in(Select company_id from Market_cap where Market_Cap 介于 0 和 10000 之间且 Date = '2017-06-30 )) 但我得到错误：

无法对包含聚合或子查询的表达式执行聚合函数。

现在编码：

 select m.date_month     
     ,sum(case when y.date_year = 2015 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then n.EBIT end) / sum(case when y.date_year = 2015 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then s.Sales end) as 'EBIT/Sales 2015'  
     ,sum(case when y.date_year = 2016 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then n.EBIT end) / sum(case when y.date_year = 2016 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then s.Sales end) as 'EBIT/Sales 2016'  
     ,sum(case when y.date_year = 2017 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then n.EBIT end) / sum(case when y.date_year = 2017 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 AND mm.Market_Cap between 0 and 10000 then s.Sales end) as 'EBIT/Sales 2017'  
 from  EBIT   as n  
     inner join  Sales   as s on  s.company_id = n.company_id  
                             and s.date_month_id = n.date_month_id  
                             and s.date_year_id = n.date_year_id  
     inner join date_year as y on y.date_year_id = n.date_year_id  
     inner join date_month as m on m.date_month_id = n.date_month_id 
     inner join Market_Cap as mm on mm.Date_Month_Id = n.Date_Month_Id
                             and mm.Company_Id = n.Company_Id
     inner join Company as c on c.Company_Id = n.Company_Id
 where y.date_year between   2015  and   2017     
     and n.EBIT<> 0  
     and s.Sales<> 0 
 group by m.date_month;

正确的输出：

Answer 1

您在此处包含的唯一代码是选择代码，而不是实际根据市值过滤的代码。你不想改变这些行吗：

inner join Market_Cap as mm on mm.Date_Month_Id = n.Date_Month_Id
                         and mm.Company_Id = n.Company_Id

将mm.Date_Month_Id 与max(Date_Month_Id) 之类的东西进行比较？

Answer 2

正确的查询（我认为）：

 select m.date_month     
     ,sum(case when y.date_year = 2015 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then n.EBIT end) / sum(case when y.date_year = 2015 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then s.Sales end) as 'EBIT/Sales 2015'  
     ,sum(case when y.date_year = 2016 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then n.EBIT end) / sum(case when y.date_year = 2016 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then s.Sales end) as 'EBIT/Sales 2016'  
     ,sum(case when y.date_year = 2017 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then n.EBIT end) / sum(case when y.date_year = 2017 AND c.Country_Id in (4,5,6) AND c.factSet_Level1_Id between 1 and 100 then s.Sales end) as 'EBIT/Sales 2017'  
 from  EBIT   as n  
     inner join  Sales   as s on  s.company_id = n.company_id  
                             and s.date_month_id = n.date_month_id  
                             and s.date_year_id = n.date_year_id  
     inner join date_year as y on y.date_year_id = n.date_year_id  
     inner join date_month as m on m.date_month_id = n.date_month_id 
     inner join Market_Cap as mm on mm.Date_Month_Id = n.Date_Month_Id
                             and mm.Company_Id = n.Company_Id
     inner join Company as c on c.Company_Id = n.Company_Id
 where y.date_year between   2015  and   2017 and c.Company_Id in (Select Company_Id from Market_Cap Where Market_Cap between 0 and 1000) 
     and n.EBIT<> 0  
     and s.Sales<> 0 
 group by m.date_month
 order by m.Date_Month asc;