我对SSAS表格具有更丰富的经验,但是在工作中,我必须支持SSAS多维多维数据集。今天早些时候,我尝试向Excel数据透视表中的SSASMD连接添加时间智能计算(并行周期)。但是无论我使用哪种语法,评估结果都会出错。
在解决错误之后,我在SQL Server Central(Stairway to MDX - Level 13: MDX Time/Date Series Functions: LastPeriods() and ParallelPeriod() Functions)上遇到了这篇文章。
我的问题:DATE维度是否必须具有类型“ TIME”才能使日期函数起作用?这是我的YOY计算的语法。
已售出的磅数实际值-v1
(
[DATES].[INVOICE_DATE].[INVOICE_DATE].CurrentMember.PrevMember
,[Measures].[Lbs Sold]
)
已售出的磅数实际值-v2
(ParallelPeriod(
[INVOICE_DATE].[DATES].CurrentMember.Level,
1,
[INVOICE_DATE].[DATES].CurrentMember
),
[Measures].[Lbs Sold])
我对DAX的经验比对MDX的经验丰富,因此任何见解将不胜感激。 我的MDX计算源...
我也很好奇,其他 Dimension Types
都有什么用?将其从“正常”更改为其他任何功能是否会使多维数据集具有其他功能?我在网上找到了一些说明(粘贴在下面),但是我很好奇这实际上是做什么的?
来自Microsoft文档...
“类型”属性设置提供有关内容的信息 服务器和客户端应用程序的维度。在某些情况下,类型 设置仅为客户端应用程序提供指导,并且 可选的。在其他情况下,例如“帐户”或“时间”维度, 维度的类型属性设置及其属性确定 基于服务器的特定行为,可能需要实施 多维数据集中的某些行为。例如, 维度可以设置为“帐户”以指示客户应用程序 标准维度包含帐户属性。
+-----------------+---------------------------------------------------------------------------+
| Type | Description |
+-----------------+---------------------------------------------------------------------------+
| Regular | Default for dimensions that are not set to a specified type |
+-----------------+---------------------------------------------------------------------------+
| Time | Used for dimensions whose attributes represent time periods |
+-----------------+---------------------------------------------------------------------------+
| Geography | Used for dimensions whose attributes represent geographical inform-ation |
+-----------------+---------------------------------------------------------------------------+
| Organization | Used for dimensions whose attributes represent organizational information |
+-----------------+---------------------------------------------------------------------------+
| BillOfMaterials | Used for dimensions whose attributes represent inventory and |
| | man-ufacturing information |
+-----------------+---------------------------------------------------------------------------+
| Accounts | Used for dimensions whose attributes represent information used for |
| | financial reporting |
+-----------------+---------------------------------------------------------------------------+
| Customers | Used for dimensions whose attributes represent information about |
| | customers |
+-----------------+---------------------------------------------------------------------------+
| Products | Used for dimensions whose attributes represent information about products |
+-----------------+---------------------------------------------------------------------------+
| Scenario | Used for dimensions whose attributes represent information about plans |
| | and strategies |
+-----------------+---------------------------------------------------------------------------+
| Quantitative | Used for dimensions whose attributes represent quantitative inform-ation |
+-----------------+---------------------------------------------------------------------------+
| Utility | Used for dimensions whose attributes represent utility information |
+-----------------+---------------------------------------------------------------------------+
| Currency | Used for dimensions whose attributes represent currency inform-ation |
+-----------------+---------------------------------------------------------------------------+
| Rates | Used for dimensions whose attributes represent currency rate inform-ation |
+-----------------+---------------------------------------------------------------------------+
| Channel | Used for dimensions whose attributes represent channel information |
+-----------------+---------------------------------------------------------------------------+
| Promotion | Used for dimensions whose attributes represent marketing pro-motion |
| | information |
+-----------------+---------------------------------------------------------------------------+
答案 0 :(得分:1)
您可以自己尝试一下。让我们选择ParallelPeriod函数:
select
ParallelPeriod(
[Customer].[Customer Geography].[Country],
3,
[Customer].[Customer Geography].[State-Province].[Hamburg]
) on 0
from [Adventure Works]
返回南澳大利亚。这没有任何意义,但是可以。因此,您的问题的答案是“ DATE维度是否必须具有类型'TIME'才能使日期函数起作用?”不是。
根据我的经验,更改维类型的唯一原因是,当您使用半累加聚合函数(即LastNonEmpty)时,它需要多维数据集内的时间维,以便确定该期间所有间隔的值集
除此之外,它只是客户端信息。客户可以使用不同类型的图标来适应其他情况。