Allow Joining of Multiple DV Datasets within a BI Publisher Data Model

Mark Ryan

We really like how DV dataset seems a lot more friendly and user intuitive when using them within a BIP Data Model.

HOWEVER:

We were trying to leverage using DV datasets in a BIP data model to isolate specific targeted data WITHOUT creating a singular all inclusive dataset.

This is due to security and aggregation issues as well as overall performance and maintenance moving forward. Not to mention our sources are from different locations at different levels of granularity. The visuals we are creating within the report really don't cross over so creating a large all inclusive dataset is pretty much overkill for what we are trying to achieve.

The issue is, we’ve always been able to join multiple objects (using Logical SQL ) within a single BIP data model and use basically a driver to join and set parameters on all of the joined tables.

The DV Datasets don’t seem to allow for this functionality (even though in this underneath it's driven by Logical SQL written against an RPD).  It doesn't easily allow you to use the same parameters to drive the child tables and it doesn't render sample data cleanly either.

This is the way we would expect it to work.  (below is using logical SQL against the RPD)

However when using DV Datasets that a pre – developed we are unable to use a singular parameter table and link it to the others (below is 2 DV Datasets that are already created)

The ‘Create Link’ is grayed out’

And if you try to drag the link over you get the following error

This is extremely unfortunate because it forces us back to using the logical sql extracted from the DV datasets and makes it manually intensive.

One other thing we noted that a calculated column within a DV data set does NOT show up in logical SQL when you try to extract it from a DV dataset. It would be nice to not have to manually put it back in when cut/paste into a BIP Datamodel.