Question

这是我的OLAP多维数据集架构

<?xml version="1.0"?>
<Schema name="mySchema">
    <Cube name="myCube">
        <Table name="fact_access_logs_views"/>
        <Dimension name="Countries" foreignKey="country_code_id">
            <Hierarchy hasAll="true" primaryKey="country_code_id">
                <Table name="dim_country_code"/>
                <Level name="CountryCodes" column="CountryCode" uniqueMembers="false"/>
            </Hierarchy>
        </Dimension>
        <Dimension name="Channels" foreignKey="channel_id">
            <Hierarchy hasAll="true" primaryKey="channel_id">
                <Table name="dim_channel"/>
                <Level name="Channels" column="shortname_chn" uniqueMembers="false"/>
            </Hierarchy>
        </Dimension>
        <Dimension name="Time" foreignKey="access_time_id">
            <Hierarchy hasAll="true" primaryKey="access_time_id">
                <Table name="dim_time_access"/>
                <Level name="Year" column="Year" uniqueMembers="false"/>
                <Level name="Month" column="Month" uniqueMembers="false"/>
                <Level name="Date" column="Date" uniqueMembers="false"/>
                <Level name="Hour" column="Hour" uniqueMembers="false"/>
            </Hierarchy>
        </Dimension>
        <Measure name="View Count" column="id" aggregator="count" formatString="#,###"/>
    </Cube>
</Schema>

我要做的是根据观看次数找出前十个国家，并在每个国家/地区进一步查看前十个频道。请帮我构建一个MDX查询。我试图遵循这个https://msdn.microsoft.com/en-us/library/ms145579 但最终在查看多维数据集时遇到错误。

Answer 1

使用TopCount和Generate，你可以这样做。

TopCount将为您提供按您选择的度量排序的最高成员。

观点排名前10位的国家：

TopCount([Countries].[CountryCode].[CountryCode],10 ,[Measures].[View Count]  )

然后，您将使用generate和currentmember迭代第一组Top Countries 并使用上面相同的逻辑但使用频道来获取顶级频道。

WITH 
  SET TOPCHANNELSperTOPCOUNTRY AS 
    Generate
    (
      TopCount
      (
        [Countries].[CountryCode].[CountryCode]
       ,10
       ,[Measures].[View Count]
      )
     ,TopCount
      (
        (
          [Countries].[CountryCode].CurrentMember
         ,[Channels].[Channels].[Channels]
        )
       ,10
       ,[Measures].[View Count]
      )
    ) 
SELECT 
  [Measures].[View Count] ON 0
 ,TOPCHANNELSperTOPCOUNTRY ON 1
FROM [myCube];

（对不起，如果某些维度名称或成员不正确）...

Answer 2

您可以使用Top Count，Generate和CrossJoin实现此目的。

IMAGE               CREATED              CREATED BY                                      SIZE                COMMENT
147465a83f44        21 seconds ago       /bin/sh -c conda install -y --file production   618.3 MB            
ac247a09a6b9        About a minute ago   /bin/sh -c #(nop) WORKDIR /pybank               0 B                 
ee7012dc6c28        About a minute ago   /bin/sh -c #(nop) ADD dir:9d26c17d067275f3836   62.45 kB            
2552d13402b6        3 hours ago          /bin/sh -c #(nop) MAINTAINER Thomas Schmelzer   0 B                 
2fd9d2e11210        4 days ago           /bin/sh -c #(nop) CMD ["/bin/bash"]             0 B                 
<missing>           4 days ago           /bin/sh -c #(nop) ENTRYPOINT ["/usr/bin/tini"   0 B                 
<missing>           4 days ago           /bin/sh -c #(nop) ENV PATH=/opt/conda/bin:/us   0 B                 
<missing>           4 days ago           /bin/sh -c apt-get install -y curl grep sed d   2.293 MB            
<missing>           4 days ago           /bin/sh -c echo 'export PATH=/opt/conda/bin:$   133 MB              
<missing>           4 days ago           /bin/sh -c apt-get update --fix-missing && ap   195 MB              
<missing>           4 days ago           /bin/sh -c #(nop) ENV LANG=C.UTF-8 LC_ALL=C.U   0 B                 
<missing>           4 days ago           /bin/sh -c #(nop) MAINTAINER Kamil Kwiek <kam   0 B                 
<missing>           3 weeks ago          /bin/sh -c #(nop) CMD ["/bin/bash"]             0 B                 
<missing>           3 weeks ago          /bin/sh -c #(nop) ADD file:76679eeb94129df23c   125.1 MB     

thomas

MDX：基于度量

2 个答案: