Axis Folding
In the consolidation phase, if there are multiple slicer axes, the Analytics Engine combines results across these axes, for each result cell. This step is known as axis folding.
Important:
Axis folding means that if a given source record has a non-null result for each slicer axis, that record is counted multiple times.
To determine whether axis folding is required, the engine considers all the filters applied to the query, from all sources: the subject area, the pivot table, and the dashboard. The net combination of these filters determines whether axis folding is needed, as follows. The following table lists the main possibilities:
| Form of Filter |
Axis Folding Performed? |
| Single member. Example: [PRODUCT].[P1].[PRODUCT CATEGORY].&[Candy] |
No |
| Single measure. Example: [MEASURES].[Units Sold] |
No |
| A tuple (combination of members or of members and a measure). Example: ([Outlet].[H1].[City].&[7],[PRODUCT].[P1].[PRODUCT CATEGORY].&[Candy]) |
No |
| Cross joins that use members wrapped in %TIMERANGE functions CROSSJOIN(%TIMERANGE([BirthD].[H1].[Date].&[10000],[BirthD].[H1].[Date].&[50000]),%TIMERANGE([BirthD].[H1].[Date].&[40000],[BirthD].[H1].[Date].&[NOW])) |
Yes |
| Other cross joins. Example: NONEMPTYCROSSJOIN([Outlet].[H1].[City].&[7],[PRODUCT].[P1].[PRODUCT CATEGORY].&[Candy]) |
No |
| The %OR function, wrapped around a set expression that lists multiple members. Example: %OR({[Product].[P1].[Product Category].&[Candy],[Product].[P1].[Product Category].&[Snack]}) |
No |
| A set expression that lists multiple members but does not use %OR. Example: {[Product].[P1].[Product Category].&[Candy],[Channel].[H1].[Channel Name].&[2]} |
Yes |
To create these expressions (as filters) in the Analyzer, you generally drag and drop items to the Filters box. To create the set expressions in the last two rows, you must use the Advanced Filter editor. Note that the engine automatically uses the %OR function when possible; the Advanced Filter editor does not display it as an option.
