Can I do a max(count(*)) in SQL?

Question

Here\'s my code:

    select yr,count(*)  from movie
join casting on casting.movieid=movie.id
join actor on casting.actorid = actor.id
where actor.name = \'Jo         


        

Answer 1

it's from this site - http://sqlzoo.net/3.htm 2 possible solutions:

with TOP 1 a ORDER BY ... DESC:

SELECT yr, COUNT(title) 
FROM actor 
JOIN casting ON actor.id=actorid
JOIN movie ON movie.id=movieid
WHERE name = 'John Travolta'
GROUP BY yr
HAVING count(title)=(SELECT TOP 1 COUNT(title) 
FROM casting 
JOIN movie ON movieid=movie.id 
JOIN actor ON actor.id=actorid
WHERE name='John Travolta'
GROUP BY yr
ORDER BY count(title) desc)

with MAX:

SELECT yr, COUNT(title) 
FROM actor  
JOIN casting ON actor.id=actorid    
JOIN movie ON movie.id=movieid
WHERE name = 'John Travolta'
GROUP BY yr
HAVING 
    count(title)=
        (SELECT MAX(A.CNT) 
            FROM (SELECT COUNT(title) AS CNT FROM actor 
                JOIN casting ON actor.id=actorid
                JOIN movie ON movie.id=movieid
                    WHERE name = 'John Travolta'
                    GROUP BY (yr)) AS A)

Answer 2

Using max with a limit will only give you the first row, but if there are two or more rows with the same number of maximum movies, then you are going to miss some data. Below is a way to do it if you have the rank() function available.

SELECT
    total_final.yr,
    total_final.num_movies
    FROM
    ( SELECT 
        total.yr, 
        total.num_movies, 
        RANK() OVER (ORDER BY num_movies desc) rnk
        FROM (
               SELECT 
                      m.yr, 
                      COUNT(*) AS num_movies
               FROM MOVIE m
               JOIN CASTING c ON c.movieid = m.id
               JOIN ACTOR a ON a.id = c.actorid
               WHERE a.name = 'John Travolta'
               GROUP BY m.yr
             ) AS total
    ) AS total_final 
   WHERE rnk = 1

Answer 3

create view sal as
select yr,count(*) as ct from
(select title,yr from movie m, actor a, casting c
where a.name='JOHN'
and a.id=c.actorid
and c.movieid=m.id)group by yr

-----VIEW CREATED-----

select yr from sal
where ct =(select max(ct) from sal)

YR 2013

Answer 4

Thanks to the last answer

SELECT yr, COUNT(title)
FROM actor
JOIN casting ON actor.id = casting.actorid
JOIN movie ON casting.movieid = movie.id
WHERE name = 'John Travolta'
GROUP BY yr HAVING COUNT(title) >= ALL
  (SELECT COUNT(title)
   FROM actor
   JOIN casting ON actor.id = casting.actorid
   JOIN movie ON casting.movieid = movie.id
   WHERE name = 'John Travolta'
   GROUP BY yr)

I had the same problem: I needed to know just the records which their count match the maximus count (it could be one or several records).

I have to learn more about "ALL clause", and this is exactly the kind of simple solution that I was looking for.

Answer 5

Use:

  SELECT m.yr, 
         COUNT(*) AS num_movies
    FROM MOVIE m
    JOIN CASTING c ON c.movieid = m.id
    JOIN ACTOR a ON a.id = c.actorid
                AND a.name = 'John Travolta'
GROUP BY m.yr
ORDER BY num_movies DESC, m.yr DESC

Ordering by num_movies DESC will put the highest values at the top of the resultset. If numerous years have the same count, the m.yr will place the most recent year at the top... until the next num_movies value changes.

Can I use a MAX(COUNT(*)) ?

---

No, you can not layer aggregate functions on top of one another in the same SELECT clause. The inner aggregate would have to be performed in a subquery. IE:

SELECT MAX(y.num)
  FROM (SELECT COUNT(*) AS num
          FROM TABLE x) y