具有嵌套聚合的 MDX 查询答案

【问题标题】：MDX query with nested aggregations具有嵌套聚合的 MDX 查询
【发布时间】：2016-05-10 08:22:59
【问题描述】：

WITH 
  MEMBER [Measures].[newCalculatedMeasure] AS 
    Sum
    (
      Filter
      (
        Descendants
        (
          [Date].[28 Days Month Calendar].CurrentMember
         ,2
         ,LEAVES
        )
       ,
          [Measures].[mymeasure] * 10
        > 
          Avg
          (
            Filter
            (
              Descendants
              (
                [Date].[28 Days Month Calendar].CurrentMember
               ,2
               ,LEAVES
              )
             ,
              [Measures].[mymeasure] > 0
            )
           ,[Measures].[mymeasure]
          )
      )
     ,[Measures].[mymeasure]
    ) 
SELECT 
  [Date].[28 Days Month Calendar].MEMBERS ON 0
 ,[Measures].[newCalculatedMeasure] ON 1
FROM [cube];

上面的查询没有评估内部平均函数，而如果我用实际平均值替换它，它的工作正常

有人可以告诉我上述措施有什么问题吗..

【问题讨论】：

标签： mdx msas

【解决方案1】：

将度量拆分为命名集是否有助于解决此上下文问题？

WITH 
  SET [MYSET] AS 
    Descendants
    (
      [Date].[28 Days Month Calendar].CurrentMember
     ,2
     ,LEAVES
    ) 
  SET [MYSETFILTERED] AS 
    Filter
    (
      [MYSET]
     ,
      [Measures].[mymeasure] > 0
    ) 
  MEMBER [Measures].[newCalculatedMeasure] AS 
    Sum
    (
      Filter
      (
        [MYSET]
       ,
        [Measures].[mymeasure] * 10 > Avg([MYSETFILTERED],[Measures].[mymeasure])
      )
     ,[Measures].[mymeasure]
    ) 
SELECT 
  [Date].[28 Days Month Calendar].MEMBERS  ON 0
 ,[Measures].[newCalculatedMeasure] ON 1
FROM [cube];

具有聚合的命名集有时会影响性能，所以如果上述情况拖慢了一切，请见谅。

【讨论】：