CubeGrid: How to handle asymmetric Facets

[jooky68]

I need to represent complex reports which have asymmetric column groupings (facets). See attached image. There are two important characteristics of the column groupings for which I am unsure how to correctly approach with SmartGWT.

1. Column sizing is not symmetrical. Notice how "% of MktVal" under "Media" is a different column width then the other "% of MktVal" columns.
2. Groupings are not symmetrical. "Core Plus 4" has different buckets then "Core Plus C"

My questions is, is there a proper way to handle this with the CubeGrid? The only approach I have been able to come up with is that I have to create unique FacetValues for each column and bucket in my report. The FacetValues that do not apply for a specific bucket must then be hidden. With reports consisting of wide and deep bucketing, this can result in a very large number of unused FacetValues. My concern, aside from programming complexity, is more of a performance concern. Is this unwarranted?

Any insight into possible approaches with the SmartGWT framework would be greatly appreciated.

[Isomorphic]

There are two approaches:

1. set hideEmptyFacetValues:true - this works by automatically hiding facet values that are not present in the data, but of course this requires all data to be loaded up front rather than using load on demand

2. use a tree facet - it can have hierarchical structure, such as the Time facet shown in the Advanced Cube sample. However, it renders as an expand-collapse tree rather than nested headers. If you need the nested header rendering instead, that's something you could sponsor.