Oracle 18c

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

BANNER	BANNER_FULL	BANNER_LEGACY	CON_ID
Oracle Database 18c Express Edition Release 18.0.0.0.0 - Production	Oracle Database 18c Express Edition Release 18.0.0.0.0 - Production Version 18.4.0.0.0	Oracle Database 18c Express Edition Release 18.0.0.0.0 - Production	0

with t (num, dt, val, days) as (
    select 1, date '2016-01-01', 0 , 0 from dual union all
    select 1, date '2016-01-02', 0 , 0 from dual union all
    select 1, date '2016-01-03', 50, 1 from dual union all
    select 1, date '2016-01-04', 50, 2 from dual union all
    select 1, date '2016-01-05', 50, 3 from dual union all
    select 1, date '2016-01-06', 0 , 0 from dual union all
    select 1, date '2016-01-07', 50, 1 from dual union all
    select 1, date '2016-01-08', 0 , 0 from dual union all
    select 1, date '2016-01-09', 0 , 0 from dual union all
    select 1, date '2016-01-10', 50, 1 from dual union all
    select 1, date '2016-01-11', 0 , 0 from dual union all
    select 1, date '2016-01-12', 0 , 0 from dual union all
    select 2, date '2016-01-01', 50, 1 from dual union all
    select 2, date '2016-01-02', 50, 2 from dual union all
    select 2, date '2016-01-03', 50, 3 from dual union all
    select 2, date '2016-01-04', 50, 4 from dual union all
    select 2, date '2016-01-05', 0 , 0 from dual union all
    select 2, date '2016-01-06', 50, 1 from dual union all
    select 2, date '2016-01-07', 0 , 0 from dual union all
    select 2, date '2016-01-08', 50, 1 from dual union all
    select 2, date '2016-01-09', 50, 2 from dual union all
    select 2, date '2016-01-10', 50, 3 from dual union all
    select 2, date '2016-01-11', 50, 4 from dual union all
    select 2, date '2016-01-12', 50, 5 from dual union all
    select 2, date '2016-01-13', 50, 6 from dual union all
    select 2, date '2016-01-14', 50, 7 from dual)
--
select num, dt, val, countup days
from t 
match_recognize (
    partition by num order by dt 
    measures count (b.val) as countup
    all rows per match
    pattern (a* b*)
    define a as val = 0, b as val > 0) 
/

NUM	DT	VAL	DAYS
1	01-JAN-16	0	0
1	02-JAN-16	0	0
1	03-JAN-16	50	1
1	04-JAN-16	50	2
1	05-JAN-16	50	3
1	06-JAN-16	0	0
1	07-JAN-16	50	1
1	08-JAN-16	0	0
1	09-JAN-16	0	0
1	10-JAN-16	50	1
1	11-JAN-16	0	0
1	12-JAN-16	0	0
2	01-JAN-16	50	1
2	02-JAN-16	50	2
2	03-JAN-16	50	3
2	04-JAN-16	50	4
2	05-JAN-16	0	0
2	06-JAN-16	50	1
2	07-JAN-16	0	0
2	08-JAN-16	50	1
2	09-JAN-16	50	2
2	10-JAN-16	50	3
2	11-JAN-16	50	4
2	12-JAN-16	50	5
2	13-JAN-16	50	6
2	14-JAN-16	50	7