![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
|
You can use : filter ( rlid.members, not isempty(sum(measures.members))) on rows If you are not selecting all the measures in the cube then you should replace measures.members with the list of selected measures. HTH, Brian "Wilbur" <Wilbur (AT) icecube (DOT) com> wrote in message news:uZMJ$6OPDHA.3408 (AT) tk2msftngp13 (DOT) phx.gbl... Brian, This works great, thanks. I was wondering though. Is there a way to use non empty function on all of the measures besides the measures.rlidkey? This way I don't get all the empty cells on the resulting cube. Thanks in advance. "Brian" <brianaltmann (AT) yahoo (DOT) com> wrote in message news:117501c33ce7$3f5a4ba0$a101280a (AT) phx (DOT) gbl... You could create a one-leve dimension, using the name as the source for Member Name and the number as the source for Member Key. Then you could write something like: with member measures.rlidkey as 'rlid.currentmember.properties("key")' select {measures.rlidkey, measures.[your measure]} on columns, rlid.members on rows from [your cube] You'd get numbers and names side by side. HTH, Brian www.geocities.com/brianaltmann/olap.html -----Original Message----- Hi Wilbur, What Tom probably meant was that you don't need TWO levels in the dimension if it is 1:1 relation. You could simply combine the members into one level instead. Hope this helps ! -----Original Message----- What would be a more appropriate design? I am hesitant to make the rlid number and rlid name different dimensions since they are really the same thing. "Tom Chester" <publicNOSPAM (AT) tomchester (DOT) net> wrote in message news:Mt%Ka.38$a_6.26576 (AT) news (DOT) uswest.net... Wilbur, the 1:1 from parent to child sounds like a questionable design. That said, to get the entire dim on rows, versus just one level, use this syntax: Measures.Members on columns, [rlid].Members on rows (assuming rlid is the dim name) tom @ the domain below www.tomchester.net "Wilbur" <Wilbur (AT) icecube (DOT) com> wrote in message news:OmAV09MPDHA.1720 (AT) TK2MSFTNGP11 (DOT) phx.gbl... I have a hirarchical dimension with two levels. The top and bottom levels have a one to one relationship so there is only one child per parent. I would like to know how to display both the parent and child as the rows and all of my measures as the columns. I am aggregating by date with a calculated member. with member LWFormStatsTime.[range] as ' Aggregate([LWFormStatstime].[all LWFormStatsTime]. [2003].[april].[20] : [LWFormStatsTime].[all LWFormStatsTime].[2003]. [april].[20]) ' select NON EMPTY {[rlid].[vch name].members} on rows, {[measures].members} on columns from LWFormStats where LWFormStatsTime. [range] This one gets the names and the measures with member LWFormStatsTime.[range] as ' Aggregate([LWFormStatstime].[all LWFormStatsTime]. [2003].[april].[20] : [LWFormStatsTime].[all LWFormStatsTime].[2003]. [april].[20]) ' select NON EMPTY {[rlid].[vch rlid].members} on rows, {[measures].members} on columns from LWFormStats where LWFormStatsTime. [range] and this one gets me the rlid and the measures. I would like to concatinate the rlid vch name members with the rlid.vch rlid members as the rows. I tried a crossjoin and found that every other row contained a rlid and every other row contained a name. The measures double up. I would like to have the measures stay the same and have an additional column come in so that I have the names in the first column and the rlids in the second column then all of the measures. Thanks for the help in advance. . . |
![]() |
| Thread Tools | |
| Display Modes | |
| |