Paul Heinrichs

This also seems to be a problem with SQL API exposing the underlying aggregation as numeric

> @pauldheinrichs Thanks for contributing! Could you please set the allocator to undefined in the branch without queryFactory as well? It should be consistent. @paveltiunov For sure, i've updated the...

@paveltiunov sorry for the ping - curious if there's any other adjustments needed from my end here to push this forward!

@igorlukanin Appreciate the response, but i believe the following part of the query `0.0` is what converts it to a floating ``` select SUM( CASE WHEN 1 = 1 THEN...

Resolves these issues 🚀 - https://github.com/cube-js/cube/issues/8153 - https://github.com/cube-js/cube/issues/5652

Tested with pre-aggregation: - Successfully created pre-aggregation SQL ``` - Executing Load Pre Aggregation SQL: scheduler-ba104a65-512d-4700-aaeb-67caa9743162 -- CREATE TABLE dev_pre_aggregations.session_funnel_main20240101_ubycvf3l_wiz2w3y_1jafait AS SELECT "session_funnel".step "session_funnel__step", date_trunc('day', ("session_funnel".t::timestamptz AT TIME ZONE 'UTC'))...

Validated non pre-aggregated queries, ``` SELECT date_trunc('day', ("session_funnel".t::timestamptz AT TIME ZONE 'UTC')) "session_funnel__time_day", CASE WHEN APPROXIMATE COUNT(DISTINCT "session_funnel".first_step_user_id) > 0 THEN 100.0 * count("session_funnel".user_id) / APPROXIMATE COUNT(DISTINCT "session_funnel".first_step_user_id) ELSE NULL...

> @pauldheinrichs Thanks for contributing! I guess `hllInit` should use https://docs.aws.amazon.com/redshift/latest/dg/r_HLL_CREATE_SKETCH.html @paveltiunov I attempted that in this https://github.com/cube-js/cube/pull/8512/commits/71e088275864cb58f2a9ed075732386e00701a3e commit actually and resulting not super succesfully when testing? Unless i tested...

Querying on on both the view and the cube break with this. @igorlukanin Just an FYI feels like a un-intended result of this change

I've done more digging - and the source of the main issue in views container `prefix: true,` it cannot find the count(*) measure