Dynamic ParallelPeriode

Hi,

with this MDX expression as a calculated member in my cube

100 / [Measures].[Sales Volume KG] * (parallelperiod([Date].[Month], 12), [Measures].[Sales Volume KG])

I get the difference of the Sales Volume from this month and this month last year. But how can I make it dynamic for drilldown in the Time Dimension, so that also year, semester, Quarters and so on are supported? I will compare this year - last year, this semester - same semester last year, this quarter - same quarter last year and so on.

Thanks

Hans

Hi Hans,

Why doesn't ParallelPeriod(Year ..) meet your needs - could you explain in the context of this Adventure Works query?

>>

select

{[Measures].[Sales Amount]} on 0,

Generate(Ascendants([Date].[Calendar].[Date].&[800]),

{[Date].[Calendar].CurrentMember,

ParallelPeriod([Date].[Calendar].[Calendar Year],

1, [Date].[Calendar].CurrentMember)}) on 1

from [Adventure Works]

-

Sales Amount
September 8, 2003 $36,027.71
September 8, 2002 $18,755.92
September 2003 $5,057,832.17
September 2002 $3,235,826.19
Q3 CY 2003 $13,670,536.57
Q3 CY 2002 $10,277,073.06
H2 CY 2003 $26,955,981.04
H2 CY 2002 $18,646,056.13
CY 2003 $41,993,729.72
CY 2002 $30,674,773.18
All Periods $109,809,274.20

>>

|||

Hi Deepak,

Thanks for your suggestions. Your code gave me the idea to specify the Dimension in full, so I name it [Date].[Periode - Year].[Year] and not only [Date].[Year] and now it works. But it works only with the [Periode - Year] Hierarchy. How can I change it, to work with all my Time hierarchies ([Date].[Periode - Year], [Date].[Periode - Week], [Date].[Periode - Reporting])? I tried

IIF( [Measures].[Sales Volume KG] = 0, 0, 100 - (100 / [Measures].[Sales Volume KG] * (parallelperiod([Date].[Periode - Year].[Year], 1) * parallelperiod([Date].[Periode - Week].[Year], 1) * parallelperiod([Date].[Periode - Reporting].[Year], 1), [Measures].[Sales Volume KG])) )

but that doesn't work. You gave me this approach just a year ago for builing YTD sums just like

IIF( sum(YTD([Date].[Periode - Year].CurrentMember) * YTD([Date].[Periode - Week].CurrentMember) * YTD([Date].[Periode - Reporting].CurrentMember),[Measures].[Sales Volume M2]) = 0, 0, sum(YTD([Date].[Periode - Year].CurrentMember) * YTD([Date].[Periode - Week].CurrentMember) * YTD([Date].[Periode - Reporting].CurrentMember),[Measures].[Commission To Market]) / sum(YTD([Date].[Periode - Year].CurrentMember) * YTD([Date].[Periode - Week].CurrentMember) * YTD([Date].[Periode - Reporting].CurrentMember),[Measures].[Sales Volume M2]))

which work fine, but this doesn't work here. How can I do it here?

Thanks

Hans

|||

Hi Hans,

With ParallelPeriod() a different approach is needed, which would depend on the attributes in the 3 hierarchies and the attribute relations (an entry in Mosha's blog shows some examples). If all the hierarchies align with the beginning of the time dimension (which isn't true for Adventure Works Calendar hierarchy), then a cube MDX script assignment like this might work - otherwise, you could describe the attributes and hierarchies in more detail:

([Measure].[SalesGrowth], [Date].[Date].Members) =

IIF( [Measures].[Sales Volume KG] = 0, 0, 100 - (100 / [Measures].[Sales Volume KG]

* (parallelperiod([Date].[Periode - Year].[Year]), [Measures].[Sales Volume KG])) );