Postgresql selecting first item of last group

Question

I have this data:

ID | Amount
1 | 0
2 | 0
3 | 0
4 | 0 ---> NULL

ID | Amount
1 | -1 ---> this row
2 | 0
3 | 0
4 | 0
5 | 0

ID | Amount
1 | 0
2 | 0
3 | -1 ---> this row
4 | 0
5 | 0
6 | 0
7 | 0

ID | Amount
1 | 0
2 | -1
3 | -1
4 | 0
5 | 0
6 | -1 ---> this row
7 | 0
8 | 0

ID | Amount
1 | 0
2 | -1
3 | 0
4 | 0
5 | 0
6 | -1 ---> this row
7 | -1
8 | 0

my logic is start at the end, and move up until we hit the first -1. However, I am unable to arrive at an SQL for this. Any help is appreciated!

Answer 1

Assuming you have -1s, you can group them by counting the 0s that follow. These can then be enumerated using dense_rank():

select id, rnk
from (select t.*, dense_rank() over (partition by grp) as rnk
      from (select t.*,
                   count(*) filter (where amount = 0) over (order by id desc) as grp
            from t
           ) t
     ) t
where amount = -1 and rnk = 1
order by id
fetch first 1 row only;

To handle your first case, you can union in a query like this:

select id, rnk
from t
where not exists (select 1 from t where amount = -1)
order by id
fetch first 1 row only