将 sql rank 转换为 mdx rank答案

【问题标题】：Convert sql rank to mdx rank将 sql rank 转换为 mdx rank
【发布时间】：2019-11-03 07:02:49
【问题描述】：

这是我的 sql 代码 - 我想将其转换为 mdx 查询。这些结果集也用于 power bi 报告。我无法将此 sql 查询写入 mdx 查询。有人可以帮帮我吗？

    Select * From(
    Select 
    dense_RANK()over(partition by t.MonthName order by t.amount desc)rank
    ,t.*
    From( 
    Select DSP.SalesPointShortName ,dd.MonthName 
    ,SUM(fs.salesAmount)Amount
      From FactSales FS
    INNER JOIN DimDate dd on fs.DateKey=dd.DateKey
    INNER JOIN DimSalesPoint DSP on DSP.SalesPointID=FS.SalesPointID
    group by dsp.SalesPointShortName ,dd.MonthName 
    )as t  
    )as f where f.rank=1

我的预期输出是：

【问题讨论】：

标签： ssas mdx

【解决方案1】：

很多谷歌搜索我从下面的链接得到答案遵循此链接后，期望结果就来了。 https://bipassion.wordpress.com/2013/06/22/mdx-dense-rank/

 WITH SET [Sorted Models] AS
    // For each month get Top 25 records, choosed a Top 25 from business case
     Generate ( 
        [Dim Date].[Month Name].[Month Name].Members,
        TopCount 
          ( nonempty(
               {
                   [Dim Date].[Month Name].CurrentMember
                  * [Dim Sales Point].[SalesPoint].[Sales Point Short Name].MEMBERS              
                }
               ,[Measures].[Sales Amount] 
                 ) 
           , 1
           ,[Measures].[Sales Amount]
          )
          )
    //Select 
    //[Measures].[Sales Amount] on 0,[Sorted Models] on 1
   //From  [MdCubeTest]
   //where [Dim Product Group Item].[Sub Category Name].&[Bag]
    MEMBER [Measures].[Rank] AS
    // Get the Rank of current member Tuple to Current Month Set
    // Do not use Sorted Models set due to dynamic for each set
    Rank ( 
        (
        [Dim Date].[Month Name].CurrentMember 
        , [Dim Sales Point].[SalesPoint].CurrentMember    
        )
  ,  Generate ( [Dim Date].[Month Name].CurrentMember,
    TopCount 
      ( nonempty(
         { [Dim Date].[Month Name].CurrentMember
          * [Dim Sales Point].[SalesPoint].[Sales Point Short Name]          
                  }
           ,[Measures].[Sales Amount] 
           ) 
       , 25
       ,[Measures].[Sales Amount]
      )
     )
    , [Measures].[Sales Amount] 
    )

MEMBER [Measures].[Previous Set Index] AS
// Get the Set Index using Rank 
( 
[Measures].[Rank] - 2 
) 
MEMBER [Measures].[Dense Rank] AS
// Get the Dense Rank using the Index position value 
CASE
 WHEN [Measures].[Rank] = 1 
  THEN 1 
 ELSE
        ( 
        [Sorted Models].Item([Measures].[Previous Set Index]), 
        [Measures].[Dense Rank] 
        ) 
     + 
        Iif 
        ( 
             ( 
             [Sorted Models].Item([Measures].[Previous Set Index]), 
             [Measures].[Sales Amount] 
             ) 
          = 
             [Measures].[Sales Amount] 
         ,0 
         ,1 
        ) 
End

SELECT
  {
      [Measures].[Sales Amount]
    , [Measures].[Rank]
    , [Measures].[Dense Rank]
    //, [Measures].[Previous Set Index]
  } ON rows
 ,NON EMPTY
     {
     // FILTER Can be used to get the TOP 3 using DENSE RANK 
      FILTER (  
               [Sorted Models]
          , [Measures].[Dense Rank] <=1
          )

     } ON columns
    FROM  [MdCubeTest]
    where [Dim Product Group Item].[Sub Category Name].&[Bag]`

【讨论】：