Return the list of tuples removing all tuples that are empty based on the indexing of the measure
This is a very fast version of pseudo - nonempty. It's worth using when applied to a huge crossjoin.
- NonEmpty is based in the indexing structure, not in fact values. Only if the measure as no NULL values in facts it will return the same result as a nonempty.
- Set has to be a crossjoin.
- Set is materialized, pay attention of the resulting size.
- If possible, a multithreaded version is used.
Special Examples (errors, null, ...)
Find here the value of the XMLA properties when not specified otherwise.