根据select中的count列返回最大行

Question

I am trying to display 1 row based off the maximum number of visits. I already have a count function in my select clause.

This is what I already have for code:

select
pat.pat_id "pat id"
,DEP.DEPARTMENT_NAME "DEPT NAME"
,COUNT (DEP.DEPARTMENT_NAME) "VISITS BY DEPT"
FROM PAT_ENC PAT
JOIN CLARITY_DEP DEP  ON PAT.DEPARTMENT_ID    = DEP.DEPARTMENT_ID
GROUP BY PAT.PAT_ID
    ,DEP.DEPARTMENT_NAME

Based on the code above here are my current results:

PAT ID  DEPT NAME           VISITS BY DEPT
Z116547 OPS PEDI REHAB OT   177
Z116547 SSH REHAB PT        85
Z116547 SSH REHAB SLP       62
Z130422 LIB WOUND CENTER    119
Z130422 WEY UROLOGY         1
Z130422 SSH AMBULATORY CARE 1

I'm trying to write code that will only display the row that has the maximum number of visits for each Pat ID.

The results I am trying to get is

PAT ID  DEPT NAME           VISITS BY DEPT
Z116547 OPS PEDI REHAB OT   177
Z130422 LIB WOUND CENTER    119

I know I need a sub query, just not sure where and how it should look. Thanks.

Answer 1

One method uses window functions:

SELECT pd.*
FROM (SELECT pat.pat_id, DEP.DEPARTMENT_NAME, 
             COUNT(*) as cnt,
             ROW_NUMBER() OVER (PARTITION BY DEP.DEPARTMENT_NAME ORDER BY COUNT(*) DESC) as seqnum
      FROM PAT_ENC PAT JOIN
           CLARITY_DEP DEP
           ON PAT.DEPARTMENT_ID = DEP.DEPARTMENT_ID
      GROUP BY PAT.PAT_ID, DEP.DEPARTMENT_NAME
     ) pd
WHERE seqnum = 1;

Note that I removed your column aliases. Having to fiddle around with escaping column names just makes it more complicated to write queries.

Answer 2

use row-number()

  select * from (select
pat.pat_id "pat id"
,DEP.DEPARTMENT_NAME "DEPT NAME"
,COUNT (DEP.DEPARTMENT_NAME) "VISITS BY DEPT",
 row_number() over(partition by pat.pat_id order by COUNT (DEP.DEPARTMENT_NAME) desc ) as rn
FROM PAT_ENC PAT
JOIN CLARITY_DEP DEP  ON PAT.DEPARTMENT_ID    = DEP.DEPARTMENT_ID
GROUP BY PAT.PAT_ID
    ,DEP.DEPARTMENT_NAME
) a where a.rn=1