Sql server 2008 MDX查询中元组内度量值的上一个值

我在报告中有以下MDX查询：

select 
({ [Measures].[Total]})   on columns,
non empty
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
having [Measures].Total>100000 on rows
from [Monthly Summary]
where ([1 Date - Month End].[Month End Date].[Month].&[20120331])

这是在3月份选择总值超过100k的行。我如何选择3月或2月总值超过100k的行意味着从上面的查询中获得结果加上上上个月超过100k的本月值的结果

现在，我在SSRS中的两个数据集中执行此操作，并在另一个数据集中执行联接和查找，从报表参数读取月份（Where子句）；这是不正确的

---

更新：无论三月的[Total]值大于100k还是二月的[Total]值大于100k，我都需要在查询结果中看到三月的总值（第二种情况下可能小于100k）

如果我了解您的要求，您需要where子句中月份的行或where子句中总数超过100K的前一个月的行。where子句将影响您在Total度量值中看到的值，因此如果您也想看到上个月的值，则不能将其保留在where子句中。要查看该行是三月还是二月，必须将其包含在以下行中：

select 
({ [Measures].[Total]})   on columns,
non empty
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
* {[1 Date - Month End].[Month End Date].[Month].&[20120228],[1 Date - Month End].[Month End Date].[Month].&[20120331]}
having [Measures].Total>100000 on rows
from [Monthly Summary]

select 
    ({ [Measures].[Total]})   on columns,
    non empty
    [Dim1].[h1].allmembers
    * [Dim1].[h2].allmembers
    * [Dim1].[h3].allmembers
    * [1 Date - Loss Date].[Date].[Year].allmembers
    * {[1 Date - Month End].[Month End Date].[Month].&[20120331].PrevMember,[1 Date - Month End].[Month End Date].[Month].&[20120331]}
    having [Measures].Total>100000 on rows
    from [Monthly Summary]

SELECT ({ [Measures].[Total]}) ON COLUMNS,
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
HAVING [Measures].[Total] > 100000
OR ([Measures].[Total], [1 Date - Month End].[Month End Date].[Month].&[20120228]) >100000 
ON ROWS
FROM [Monthly Summary]
WHERE ([1 Date - Month End].[Month End Date].[Month].&[20120331])

您是否从参数中获取月份，并因此寻找获取前一个月的通用方法？如果是这样，您可以这样做：

select 
({ [Measures].[Total]})   on columns,
non empty
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
* {[1 Date - Month End].[Month End Date].[Month].&[20120228],[1 Date - Month End].[Month End Date].[Month].&[20120331]}
having [Measures].Total>100000 on rows
from [Monthly Summary]

select 
    ({ [Measures].[Total]})   on columns,
    non empty
    [Dim1].[h1].allmembers
    * [Dim1].[h2].allmembers
    * [Dim1].[h3].allmembers
    * [1 Date - Loss Date].[Date].[Year].allmembers
    * {[1 Date - Month End].[Month End Date].[Month].&[20120331].PrevMember,[1 Date - Month End].[Month End Date].[Month].&[20120331]}
    having [Measures].Total>100000 on rows
    from [Monthly Summary]

SELECT ({ [Measures].[Total]}) ON COLUMNS,
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
HAVING [Measures].[Total] > 100000
OR ([Measures].[Total], [1 Date - Month End].[Month End Date].[Month].&[20120228]) >100000 
ON ROWS
FROM [Monthly Summary]
WHERE ([1 Date - Month End].[Month End Date].[Month].&[20120331])

更新： 若要在当前月份或上个月大于100000时获取当前值，无论当前月份的总数如何，都可以执行以下操作：

select 
    ({ [Measures].[Total]})   on columns,
    non empty
    [Dim1].[h1].allmembers
    * [Dim1].[h2].allmembers
    * [Dim1].[h3].allmembers
    * [1 Date - Loss Date].[Date].[Year].allmembers

    having [Measures].Total>100000 or ([1 Date - Month End].[Month End Date].[Month].&[20120331].PrevMember,  [Measures].Total>100000) on rows
    from [Monthly Summary] 
    where [1 Date - Month End].[Month End Date].[Month].&[20120331]

---

在HAVING子句中，WHERE子句与度量值相结合，以筛选出不符合指定月份（2012年3月）标准的任何行。但它不会对OR之后的表达式执行此操作，因为带有PrevMember的元组覆盖WHERE子句

您可以使用以下内容：

select 
({ [Measures].[Total]})   on columns,
non empty
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
* {[1 Date - Month End].[Month End Date].[Month].&[20120228],[1 Date - Month End].[Month End Date].[Month].&[20120331]}
having [Measures].Total>100000 on rows
from [Monthly Summary]

select 
    ({ [Measures].[Total]})   on columns,
    non empty
    [Dim1].[h1].allmembers
    * [Dim1].[h2].allmembers
    * [Dim1].[h3].allmembers
    * [1 Date - Loss Date].[Date].[Year].allmembers
    * {[1 Date - Month End].[Month End Date].[Month].&[20120331].PrevMember,[1 Date - Month End].[Month End Date].[Month].&[20120331]}
    having [Measures].Total>100000 on rows
    from [Monthly Summary]

SELECT ({ [Measures].[Total]}) ON COLUMNS,
[Dim1].[h1].allmembers
* [Dim1].[h2].allmembers
* [Dim1].[h3].allmembers
* [1 Date - Loss Date].[Date].[Year].allmembers
HAVING [Measures].[Total] > 100000
OR ([Measures].[Total], [1 Date - Month End].[Month End Date].[Month].&[20120228]) >100000 
ON ROWS
FROM [Monthly Summary]
WHERE ([1 Date - Month End].[Month End Date].[Month].&[20120331])

---

行上不需要非emtpy，因为

null<100000

谢谢您的回答。对不起，我没有明确说明我的要求。如果“先前”[total]>100k，我仍然需要查看“当前”[total]（可能小于100k）。此查询提供的当前月或上个月的总值超过100k。@FrankGoortani我更新了我的查询，以便显示3月份的总值（即当前总值），感谢Stacia的回答。对不起，我没有明确说明我的要求。如果“先前”[total]>100k，我仍然需要查看“当前”[total]（可能小于100k）。此查询提供了本月或上个月超过100k的总值。根据我目前对贵方要求的了解，我修改了我的答案。