Can anyone help me in conversting below oracle query part into Teradata.
COUNT (DISTINCT ( case when column1 in ('aaa' , 'bbb') and column2 is null then column3
end )) OVER ( partition by some_col, some_col)
Thanks so much.
A quick look at the code, I can see that when you remove distinct, it works. Now since you may land up without distinct( i.e duplicates), then you can think of putting this in a derived query and then take only the distinct results.
You can verify with data.