I have one series as '12345234123123543'. How to write a function in SQL to find out which digit has most occurrence in a given series?

Question

I have one series as '12345234123123543'. How to write a function in SQL to find out which digit has most occurrence in a given series?

Answer 1

If you are using SQL Server then You may do like this

DECLARE @Str VARCHAR(MAX)='12345234123123543'

;WITH CTE AS(
     SELECT CAST('' AS NVARCHAR(1)) AS CH, 1 AS CNT
     UNION ALL
     SELECT CAST(SUBSTRING(@Str,CNT,1)AS NVARCHAR(1)) AS CH , CNT+1
     FROM CTE 
     WHERE CNT<=LEN(@Str)
)
SELECT TOP 1 CH,COUNT(CH) 
FROM CTE
WHERE CH <> '' 
GROUP BY CH 
ORDER BY COUNT(CH) DESC

Answer 2

If you are using Postgres you can use the following:

select c, count(*) as cnt
from unnest(string_to_array('12345234123123543',null)) as t(c)
group by c
order by cnt desc
limit 1;

Output:

c | cnt
--+----
3 |   5

---

Edit: inspired by Jaydip's answer, this can indeed be expressed with standard SQL only:

with recursive cte (ch, cnt) as (
     values ('', 1)
     union all
     select substring('12345234123123543' from cnt for 1) as ch, cnt+1
     from cte 
     where cnt <= character_length('12345234123123543')
)
select ch, count(ch) 
from cte
where ch <> '' 
group by ch 
order by count(ch) desc
fetch first 1 rows only