@User thank you! Was looking at it earlier, but still unclear where sum is calculated :) From your perspective, how hard will it be to add dynamic aggregations (sums, counts, min, ...). Aggregation spec can be extended easily, how about actual calculations?

I don't think it should be very hard. @Mayank ^^ wdyt?

Definitely doable, but may require some work. For example, we may want have multiple aggregation functions on the same column, which would mean we would need to introduce some sort of derived column, and ensure that the query execution layer understands that. I think @Jackie may have already done this for StarTree, so there may be some work to leverage from there.

That what I thought as well. Aggregations spec can be extended with derived columns (name, source column, function)

There is already a concept of derived columns in schema