On Tue, 16 Oct 2018 at 19:30, Anthony Sotolongo <[email protected]> wrote:
>
> Hola Jorge, puedes utilizar las funciones
> ventanas(https://www.postgresql.org/docs/10/static/functions-window.html),
> algo como esto, debes ajustarlo a tu escenario:
>
> with a as (
> select row_number() over ()as id, id_tip_mov,mto_can, case id_tip_mov
> when 2 then 0-mto_can
> else
> mto_can
> end as valor
> from (select * from (VALUES (1, 100), (1, 300),(2, 6),(2, 4),(1,7)) AS t
> (id_tip_mov,mto_can)) sub)
>
> select id_tip_mov,mto_can,sum (valor) over (order by id rows between
> unbounded preceding and current row) from a
>
me parece que te estas complicando de mas, el WITH no lo necesitas.
deberÃa bastar con:
select fec_emi, id_tip_mov, mto_can,
sum(case when id_tip_mov = 1 then mto_can else mto_can*(-1) end)
over (order by fec_emi rows between unbounded
preceding and current row) mto_tot
from mov;
--
Jaime Casanova www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
