Payment fully received - sql

Question

We are raising bills to clients on various dates, and the payment is received in a irregular way. We need to calculate the payment delay days till full payments are received for a particular payment due. The data is sample data for only one client 0123
table Due (id, fil varchar(12), amount numeric(10, 2), date DATE)
table Received (id, fil varchar(12), amount numeric(10, 2), date DATE)
Table Due:
id fil amount date
----------------------------------------
1. 0123 1000. 2019-jan-01
2. 0123 1500 2019-jan-15
3. 0123 1200. 2019-jan-25
4. 0123 1800. 2019-feb-10
Table Received:
id. fil. amount. date
-----------------------------------------
1. 0123 1000. 2019-jan-10
2. 0123 500. 2019-jan-20
3. 0123 1300. 2019-jan-25
4. 0123 400. 2019-feb-08
5. 0123 1000. 2019-feb-20
The joined table should show:
fil. due_date due_amount. received_amount date delay
------------------------------------------------------------------------
0123 2019-jan-01 1000. 1000 9
0123 2019-jan-15. 1500. 500
0123 1300. 10(since payment completed on 25th jan)
0123 2019-jan-25. 1200. 400.
0123 1000. 26
0123 2019-feb-10. 1800.
I have tried to be as accurate as possible in calculations......Please excuse if there is some in advertant error. I was just coming around to writing a script to do this, but maybe someone will be able to suggest a proper join.
Thanks for trying..

Answer 1

As #DavidHempy said, this is not possible without knowing for which invoice each payment is meant. You can calculate how many days it's been since the account was at 0, which might help:
with all_activity as (
select due.date,
-1 * amount as amount
from due
union all
select received.date,
amount
from received),
totals as (
select date,
amount,
sum(amount) over (order by date),
case when sum(amount) over (order by date) >=0
then true
else false
end as nothing_owed
from all_activity)
select date,
amount,
sum,
date - max(date) filter (where nothing_owed = true) OVER (order by date)
as days_since_positive
from totals order by 1
;
date | amount | sum | days_since_positive
------------+----------+----------+---------------------
2019-01-01 | -1000.00 | -1000.00 |
2019-01-10 | 1000.00 | 0.00 | 0
2019-01-15 | -1500.00 | -1500.00 | 5
2019-01-20 | 500.00 | -1000.00 | 10
2019-01-25 | -1200.00 | -900.00 | 15
2019-01-25 | 1300.00 | -900.00 | 15
2019-02-08 | 400.00 | -500.00 | 29
2019-02-10 | -1800.00 | -2300.00 | 31
2019-02-20 | 1000.00 | -1300.00 | 41
(9 rows)
You could extend this logic to figure out the last due date from which they were above 0.