This section contains information that further explains how dimension group data is displayed.
Dimension groups enable the user to select values from each of the dimensions contained in them. If the relationships made by a dimension group were instead created with hierarchy, once a value had been selected from one of the branches, then the remaining dimension values would no longer be valid for refinement.
For example, in mutual funds data, a user may want to navigate on a variety of performance criteria. A Performance dimension group that contains the YTD Total Returns, 1 Year Total Returns, and Five Year Total Returns dimensions would enable the user to select criteria from all three dimensions. If the same relationship had been created using dimension hierarchy, then once a selection had been made from the 1 Year Total Returns branch, the other two branches would no longer be available for navigation.
The display order of dimension groups is determined by the ranking of the individual dimensions within the groups. A dimension group inherits the highest rank of its member dimensions. For example, if the highest-ranked dimension in dimension group A has a rank of 5, and the highest-ranked dimension in group B has a rank of 7, then group B will be ordered before group A.
Dimension groups are also ranked relative to dimensions not within explicit groups. Continuing the previous example, an implicit dimension with a rank of 6 would be ordered after dimension group B, but before group A.
Dimensions with the same rank are ordered by name. It is important to note that dimension name, not dimension group name, determines the display order in this situation: Dimension groups are ordered according to their highest alphanumerically-ranked member dimensions. Therefore, dimension group Z, which contains dimension H, will be ordered before dimension group A, which contains dimension I.
For more information about ranking, see the Developer Studio online help.