I am trying to put together my first query for a pentaho CDE dashboard chart.
Starting Query
WITH
SET [~COLUMNS] AS
{[DimProgram.Name].[Name].MEMBERS}
SET [~ROWS] AS
{[DimTime.CalendarYearMonth].[CalendarYearMonth].MEMBERS}
SELECT
NON EMPTY
CrossJoin
(
[~COLUMNS]
,{[Measures].[SubmissionCount]}
) ON COLUMNS
,NON EMPTY
[~ROWS] ON ROWS
FROM [PSE_FactSubmission];
This query returns the data I want but needs to be tweaked a bit to be ready for actual use. I want to sort by date descending and limit to only the past 12 months.
I've read several webpages on sorting in MDX, but haven't been able to put together a query that will run. When the query doesn't run just an "Error" prompt.
Ordering Attempt
WITH
SET [~COLUMNS] AS
{[DimProgram.Name].[Name].MEMBERS}
SELECT
NON EMPTY
CrossJoin
(
[~COLUMNS]
,{[Measures].[SubmissionCount]}
) ON COLUMNS
,NON EMPTY
Order
(
[DimTime.CalendarYearMonth].[CalendarYearMonth].MEMBERS
,[DimTime.CalendarYearMonth].CurrentMember.Member_Key
,DESC
) ON ROWS
FROM [PSE_FactSubmission];
Any tips on sorting or how to limit to the past X months would be very appreciated.
Usually a Date/Time dimension is ordered naturally in the cube design so there is no need to use Order. I don't need to with the cubes I use.
If it is in a strange order in the cube then you need to break (B) this hierarchical order by using BASC or BDESC:
WITH
SET [~COLUMNS] AS
{[DimProgram.Name].[Name].MEMBERS}
MEMBER [Measures].[orderMeas] AS
[DimTime.CalendarYearMonth].CurrentMember.Member_Key
SET [~ROWS] AS
Order
(
[DimTime.CalendarYearMonth].[CalendarYearMonth].MEMBERS
,[Measures].[orderMeas]
,BASC
)
SELECT
NON EMPTY
CrossJoin
(
[~COLUMNS]
,{[Measures].[SubmissionCount]}
) ON COLUMNS
,NON EMPTY
[~ROWS] ON ROWS
FROM [PSE_FactSubmission];
To get the most recent 12 months you can use the Tail function - better to use it against NonEmpty months:
WITH
SET [~COLUMNS] AS
{[DimProgram.Name].[Name].MEMBERS}
MEMBER [Measures].[orderMeas] AS
[DimTime.CalendarYearMonth].CurrentMember.Member_Key
SET [~ROWS] AS
Order
(
NonEmpty
(
[DimTime.CalendarYearMonth].[CalendarYearMonth].MEMBERS
,[Measures].[SubmissionCount]
)
,[Measures].[orderMeas]
,BASC
)
SELECT
NON EMPTY
CrossJoin
(
[~COLUMNS]
,{[Measures].[SubmissionCount]}
) ON COLUMNS
,NON EMPTY
Tail
(
[~ROWS]
,12
) ON ROWS
FROM [PSE_FactSubmission];
Hi Andrew - against AdvWrks I've got the following running without any errors. I needed to change Member_Key to MemberValue:
WITH
SET [~COLUMNS] AS
[Product].[Product Categories].[Product]
MEMBER [Measures].[orderMeas] AS
[Date].[Calendar].CurrentMember.MemberValue
SET [~ROWS] AS
Order
(
NonEmpty
(
[Date].[Calendar].[Month].MEMBERS
,[Measures].[Internet Sales Amount]
)
,[Measures].[orderMeas]
,ASC
)
SELECT
NON EMPTY
CrossJoin
(
[~COLUMNS]
,{[Measures].[Internet Sales Amount]}
) ON COLUMNS
,NON EMPTY
Tail
(
[~ROWS]
,12
) ON ROWS
FROM [Adventure Works];
Related
Working pentaho dashboard trying to generate report to get transaction count per day/Week/Month between two date.
Below MDX Query return transaction count 1 because 'created' is not group by
WITH
SET [~COLUMNS] AS Filter([created].DefaultMember.Children, [created].CurrentMember.name >= "2014-10-01" AND Left([created].CurrentMember.Name, 10) <= "2014-10-02")
SET [~ROWS] AS
{[markup].[markup].Members}
SELECT
NON EMPTY CrossJoin([~COLUMNS], {[Measures].[Transaction Count]}) ON COLUMNS,
NON EMPTY [~ROWS] ON ROWS
FROM [transaction]
How to i can group column '[created]'
Why not just move the measure into the WHERE clause:
WITH
SET [~COLUMNS] AS Filter([created].DefaultMember.Children, [created].CurrentMember.name >= "2014-10-01" AND Left([created].CurrentMember.Name, 10) <= "2014-10-02")
SET [~ROWS] AS
{[markup].[markup].Members}
SELECT
NON EMPTY [~COLUMNS] ON COLUMNS,
NON EMPTY [~ROWS] ON ROWS
FROM [transaction]
WHERE [Measures].[Transaction Count]
I am need to delete all 2 month from set. This code return all date range without excepting 2 month.
SELECT
{[Measures].[In]} ON COLUMNS,
EXCEPT([Date].[2014].[1] : [Date].[2016].[2], [Date].[Month].[2]) ON ROWS
FROM [Shop hourly]
Print screen for whytheq
My decision based on whytheq answear. I create a dimension for all kind of dates, and except them. Example:
SELECT {[Measures].[In sum]} ON COLUMNS,
NON EMPTY
{[Shop].[11], [Shop].[22], [Shop].[33]} *
Except([Quarter].Children, [Quarter].[2]) *
[Month].Children ON ROWS
FROM [Shop hourly]
WHERE
([Date].[2013].[1].[1] : [Date].[2016].[1].[1]) *
Except([Year].Children, [Year].[2014])
In AdventureWorks I can do the following:
SELECT
[Measures].[Internet Sales Amount] ON 0
,NON EMPTY
Except
(
[Date].[Calendar].[Month].&[2005]&[7]
:
[Date].[Calendar].[Month].&[2008]&[7]
,Exists
(
[Date].[Calendar].[Month].MEMBERS
,[Date].[Calendar Quarter of Year].&[CY Q3]
)
) ON 1
FROM [Adventure Works];
So adapting the above to your cube maybe looks like the following:
SELECT
{[Measures].[In]} ON COLUMNS
,Except
(
[Date].[2014].[1] : [Date].[2016].[2]
,Exists
(
[Date].MEMBERS
,[Date].[Month].[2]
)
) ON ROWS
FROM [Shop hourly];
WITH
MEMBER [Measures].[ParameterCaption] AS [Estate].[Week].CURRENTMEMBER.MEMBER_CAPTION
MEMBER [Measures].[ParameterValue] AS [Estate].[Week].CURRENTMEMBER.UNIQUENAME
MEMBER [Measures].[ParameterLevel] AS [Estate].[Week].CURRENTMEMBER.LEVEL.ORDINAL
SELECT
{[Measures].[ParameterCaption], [Measures].[ParameterValue], [Measures].[ParameterLevel]} ON COLUMNS
,
NON EMPTY (
ORDER (
EXCEPT( [Estate].[Week].[ALL].CHILDREN
, { [Estate].[Week]})
, ( [Estate].[Week].MEMBERVALUE)
, ASC
)
) ON ROWS
From [EstateRpt]
WHERE Filter([V Estate Weekly Rpt].[Week].CHILDREN, [V Estate Weekly Rpt].[Week].MEMBERVALUE = 'NONE')
Hi, i am new to the MDX. I want to filter the week which is not equal to "NONE"? by default, week is set the "NONE", so it will appear the NONE data in cube. I want to filter this NONE.
I do try the WHERE clause but it show the error to me which i do not figure out what is the problem
If you want to filter out 'NONE' then you need to use this in your filter:
<> 'NONE'
Also you need to use CurrentMember within your filter like this example:
SELECT
{
[Date].[Calendar].[Month].&[2006]&[7]
,[Date].[Calendar].[Date].&[20050220]
}
*
{[Measures].[Reseller Sales Amount]} ON COLUMNS
,
{[Product].[Category].Children}
*
{[Geography].[Geography].[Country].MEMBERS} ON ROWS
FROM [Adventure Works]
WHERE
Filter
(
[Geography].[State-Province].MEMBERS
,
[Geography].[State-Province].CurrentMember.Member_Caption <> 'California'
);
I have the following MDX:
SELECT
NON EMPTY
{
[Measures].[Returns]
,[Measures].[Returns]
} ON COLUMNS
,NON EMPTY
{[Employees].[Company].[Company].ALLMEMBERS}
DIMENSION PROPERTIES
MEMBER_CAPTION
,MEMBER_UNIQUE_NAME
ON ROWS
FROM
(
SELECT
StrToSet
("[Exec Date].[Hierarchy].[Month Num].&[2014]&[1]"
,CONSTRAINED
) ON COLUMNS
FROM [cbSales]
)
WHERE
IIF
(
StrToSet
("[Exec Date].[Hierarchy].[Month Num].&[2014]&[1]"
,CONSTRAINED
).Count
= 1
,StrToSet
("[Exec Date].[Hierarchy].[Month Num].&[2014]&[1]"
,CONSTRAINED
)
,[Exec Date].[Hierarchy].CurrentMember
);
Both columns shows returned data for month 2014/01
I want to have data for the next month (2014/02) in the second column.
How can I change the query to achieve my goal.
I'll try to replicate with AdvWks cube, so currently not tested:
WITH
MEMBER [Measures].[ReturnsNextMth] AS
(
[Measures].[Returns]
,[Exec Date].[Hierarchy].CurrentMember.NextMember
)
SELECT
NON EMPTY
{
[Measures].[Returns]
,[Measures].[ReturnsNextMth]
} ON COLUMNS
,NON EMPTY
{
[Employees].[Company].[Company].ALLMEMBERS
*
[Exec Date].[Hierarchy].[Month Num].ALLMEMBERS
} ON ROWS
FROM
(
SELECT
StrToSet
("[Exec Date].[Hierarchy].[Month Num].&[2014]&[1]"
,CONSTRAINED
) ON COLUMNS
FROM [cbSales]
);
Something similar in AdvWks:
WITH
MEMBER [Measures].[ReturnsNextMth] AS
(
[Measures].[Internet Order Count]
,[Date].[Calendar].CurrentMember.NextMember
)
SELECT
NON EMPTY
{
[Measures].[Internet Order Count]
,[Measures].[ReturnsNextMth]
} ON COLUMNS
,NON EMPTY
{
[Sales Territory].[Sales Territory].[Country].MEMBERS
*
[Date].[Calendar].[Month].ALLMEMBERS
} ON ROWS
FROM
(
SELECT
[Date].[Calendar].[Month].&[2008]&[1] ON COLUMNS
FROM [Adventure Works]
);
This gives these results. I found this quite surprising as the sub-select seems to suggest we only have Jan-2008 available, but sub-select is a bit of a strange creature. This article explains: http://bisherryli.com/2013/02/08/mdx-25-slicer-or-sub-cube/. So our measure [ReturnsNextMth] still functions ok.
Hi,
I want to order the PriceValues field in the following query in Ascending Order
SELECT
NON EMPTY
Hierarchize
(
Union
(
{[PriceValues].[Grand Total]}
,[PriceValues].[Grand Total].Children
)
) ON COLUMNS
,NON EMPTY
Hierarchize
(
Union
(
{[Offer].[Grand Total]}
,[Offer].[Grand Total].Children
)
) ON ROWS
FROM [Act_RenewDetails]
WHERE
{[Measures].[Revenue]};
To order a measure in ascending numerical order:
{ORDER({[PriceValues].[Grand Total],[PriceValues].[Grand Total].Children}, [Measures].[Revenue], BASC)} ON COLUMNS
Edit: Or you can order alphabetically by the member names of a dimension:
{ORDER({[PriceValues].[Grand Total],[PriceValues].[Grand Total].Children}, [PriceValues].CurrentMember.name, BASC)} ON COLUMNS
(I think .name might be .memberName or even .caption - my memory is rusty.)