Javadoc
Creates a DataQueryGroups object. It is mandatory to group the queries by
the following criteria: 1) partition / year 2) organisation unit level
3) period type 4) aggregation type. The DataQueryGroups contains groups of
queries. The query groups should be run in sequence while the queries within
each group should be run in parallel for optimal performance. Currently
queries with different
AnalyticsAggregationType are run in sequence.
If the number of queries produced by this grouping is equal or
larger than the number of optimal queries, those queries are returned. If
not it will split on the data element dimension, data set dimension and
organisation unit dimension, and return immediately after each step if
optimal queries are met.
It does not attempt to split on period dimension as splitting on columns
with low cardinality typically does not improve performance.