我有一张桌子
| Location | Unit | ...
+----------|------+----
| A | 1 | ...
| A | 1 | ...
| B | 1 | ...
| A | 2 | ...
| C | 2 | ...
| ... | ... | ...
我希望计算一个新表,其中包含每个单位的“平均”位置值,具体如下:
| Location | Unit | Weight |
+----------|------+--------+
| A | 1 | 0.6667 |
| B | 1 | 0.3333 |
| A | 2 | 0.5 |
| C | 2 | 0.5 |
| ... | ... | ... |
当然,获得总数
非常简单select unit, location, count(*)
from table1
group by unit, location;
并创建表
create table table2 (Unit nvarchar(50), Location int, Weight float);
但我不确定如何用平均数据填充它。 (这并不难,但不知怎的,我被卡住了......自从我在SQL Server上工作已经很多年了。)
答案 0 :(得分:5)
您可以使用COUNT OVER
:
select distinct
location, unit,
cast(count(*) over (partition by unit, location) as decimal) /
cast(count(*) over (partition by unit) as decimal) as weight
from mytable
order by unit, location;
答案 1 :(得分:2)
您想要加权平均值,单位为权重。您需要每个位置/单元组合的总和以及每个location
的总和。要产生你的输出:
select location, unit, (1.0*unit) / sum(unit) over (partition by location)
from table1
group by location, unit;
如果1.0*
实际上是一个整数,unit
只是转换为十进制值的快捷方式。
编辑:
如果您只想要计数,那么这应该有效:
select location, unit, count(*) as cnt,
(1.0*count(*)) / sum(count(*)) over (partition by unit)
from table1
group by location, unit;
答案 2 :(得分:0)
虽然你还没有说过有一个重量栏,但我假设必须有(否则,你的体重是什么意思?)
select unit, location, AVG(weight)
from table1
group by unit, location;
作为选择语句
select unit, location, AVG(weight)
into table2
from table1
group by unit, location;
作为插入新表