Question

我似乎很难解决什么问题。数据如下表所示

表格1

num_users | time_taken
----------------------
10.1      | 11.2
10.1      | 11.5
10.2      | 11.1
10.4      | 11.9
11.01     | 12.2
11.1      | 11.8
11.7      | 13.2
11.72     | 13.1

我想计算平均值。每个time_taken值的num_users四舍五入为0十进制。这是我尝试过的。

select 
 round(num_users,0) as num_users
 , round(avg(time_taken),1) as time_taken
from table1 as a
group by a.num_users
order by a.num_users

然而，这为每个不同的num_users值提供了多行（四舍五入为0十进制） - 这里有任何想法吗？

预期产出

  num_users | time_taken
    ----------------------
    10      | 11.4
    11      | 12.0
    12      | 13.2

Answer 1

在group by：

中重复表达式

select round(num_users, 0) as num_users,
       round(avg(time_taken), 1) as time_taken
from table1 as a
group by round(num_users, 0)
order by round(num_users, 0);

我说要重复它，因为表达式的名称与列的名称相同。或者，你可以写：

select round(num_users, 0) as new_num_users,
       round(avg(time_taken), 1) as time_taken
from table1 as a
group by new_num_users
order by new_num_users;

甚至：

group by 1
order by 1

即使您从group by a.num_users中删除限定符，Postgres仍会将其解释为列名而不是新别名。

SQL - “group by”没有给出预期的结果

1 个答案: