Vous pouvez compter le nombre de valeurs, trouver les plus fréquentes, et également filtrer en fonction du nombre de valeurs :

select x.* from (select val, count(*) as cnt, row_number() over (order by count(*) desc ) as seqnum, count(*) over () as num_vals count(*) over (partition by count(*)) as cnt_cnt from table group by val ) x where cnt_cnt = 1 and seqnum = 1 and num_vals > 1;

En fait, vous pouvez utiliserhaving子句和order bypour réaliser :

select val from (select val, count(*) as cnt, count(*) over () as num_values from table group by val ) v where num_values > 1 order by cnt desc;