I need to walk a date column partitioned by the value of another column (UID) to find the earliest date for that group.
I do not want to collapse the rows.
I've tried a number of approaches to this with a SQL perspective, but I'm starting to think Magic ETL wants me to use a Group by
.
Here are a couple of examples of my experimentation to illustrate what I'm trying to do.
min( active_date ) over( partition by acct_id )
first_non_null_value( min( active_date ) over( partition by acct_id order by active_date ASC ) )
I could group then join to get the value, I was just trying to optimize the flow.