加入并聚合两列 - 每个月甚至几个月没有数据？答案

【问题标题】：Join and aggregate on two columns - for every month even months with no data?加入并聚合两列 - 每个月甚至几个月没有数据？
【发布时间】：2012-10-17 22:36:59
【问题描述】：

使用 SQL Server 2005。

我有一张包含日历月份的表格

Month,  fiscalorder
june,1
july,2
..
may,12

还有一张有员工的桌子和每月重复的金额

employee, month, amount
john, july, 10
john, july, 3
john, august,2
mary, june, 2
mary, feb, 5

我需要按月加入和汇总这些数据，但每个月（甚至没有数据的月份）都要为每个员工报告，然后是员工的财务订单。

输出：

june, john, 0
july, john, 13
august,john,2
sept, john, 0
..
june,mary,2

【问题讨论】：

标签： tsql join aggregate monthcalendar

【解决方案1】：

假设 Sql Server 2005+

Declare @CalenderMonths Table ([Month] Varchar(20),FiscalOrder Int)

Insert Into @CalenderMonths Values
('June',1),('July',2),('August',3),('September',4),('October',5),('November',6),
('December',7),('January',8),('February',9),('March',10),('April',11),('May', 12)

Declare @Employee Table(employee varchar(50), [month] Varchar(20), amount int )
Insert Into @Employee Values('john', 'July', 10),('john', 'July',3),('john','August',2),('mary','June',2),('mary', 'February',5)

;with cte as
(
    Select employee,[month],TotalAmount = sum(amount)
    from @Employee
    group by employee,[month]
)

select x.[Month],x.employee,amount = coalesce(c.TotalAmount,0)
from (
select distinct c.[Month],e.employee
from @CalenderMonths c cross join cte e)x
left join cte c on x.[Month] = c.[Month] and x.employee = c.employee
order by 2

【讨论】：

【解决方案2】：

SELECT month,employee,SUM(amount) amount
FROM(
  SELECT m.month, e.employee, ISNULL(s.amount, 0) AS amount
  FROM dbo.months AS m
  CROSS JOIN (SELECT DISTINCT employee FROM dbo.sales) AS e
  LEFT JOIN dbo.sales AS s
  ON s.employee = e.employee
  AND m.month = s.month
)X
GROUP BY month, employee

【讨论】：

这不是聚合是吗？
不，不是。不知何故错过了这个要求。更新了上面的代码以包含此要求。