SQL DB2计算模式

Question

我使用DB2和有两个表：

用户：

• ID - 关键
• 教育（整数值表示级）
• 年龄

付款方式：

• 店
• 总和
• ID - 外键

我要计算教育的统计模式，每家商店，现在我想类似的东西： 1）计算店X模式：

select u.education    
    from users u, payments p 
    where u.id = p.id AND    
    p.shop = 'X'  
    group by u.education order by count(*) desc 
    fetch first 1 rows only;      

此查询工作正常

2）计算教育的模式，每店：

select p.Shop as Shop,      
    avg(u.age) as AvgAge,          
    (select u1.education           
     from users u1          
     where u1.id = p.id         
     group by u1.education           
     order by count(*) desc           
     fetch first 1 rows only) as ModeEdu        
    from users u, payments p      
    where u.id = p.id          
    group by p.Shop;       

这个查询给出了一个错误：

SQLCODE = -119，ERROR：确定了具有 子句中的列或表达式无效

Answer 1

在这里你去。我已经更改为“现代”连接语法。你20年前使用的语法是“老”。我强烈推荐使用新的语法。

WITH eduCount AS 
(
    select u1.id, u1.education, count(*) as c 
    from users u1 
    group by u1.id u1.education           
), byUser AS 
( 
    select id, education, c, 
    ROW_NUMBER() OVER (PARTITION BY id, education ORDER BY c DESC) AS RN 
    from eduCount 
) 
select p.Shop as Shop,      
    avg(u.age) as AvgAge,          
    max(byUser.education) as ModeEdu 
from users u 
join payments p on u.id = p.id 
left join byUser on byUser.id = u1.id AND rn = 1 
group by p.Shop;