Hi, After banging my head for a few hours trying to implement an array aggregation UDF (like Postgres’ array_agg()), I realized that there’s no way of keeping non-primitive values in the @Workspace of a UDF.
I posted a wish on JIRA [1], but thought I’d also ask here if there’s any workaround to this limitation. Thanks, [1] https://issues.apache.org/jira/browse/DRILL-5421 — Manuel Aristarán <[email protected]> Datawheel LLC Cambridge, MA
