我正在尝试跟踪使用前一年的数据计算的滚动金额。
目前我将此作为我的SQL:
SELECT lngTIMEID as Id,
lngEMPID as EmployeeId,
dtmdateapp AS [Date],
stroccur AS [Value],
(SELECT Sum(stroccur)
FROM [Attendance].[dbo].timeuse a
WHERE a.dtmdateapp between DateAdd(d, -366, d.dtmdateapp)
AND d.dtmDATEAPP
AND a.lngempid = d.lngempid) AS Total
FROM [Attendance].[dbo].[timeuse] d
WHERE lngEMPID = 1844140 AND absence = 'Unscheduled' AND lngRID IN (1,2,3,4,5,7,8,9,10,11,12,13,14)
这会带来这样的结果:
+--------+------------+-----------+-------+-------+
| Id | EmployeeId | Date | Value | Total |
+--------+------------+-----------+-------+-------+
| 330435 | 1844140 | 4/17/2017 | 0.25 | 0.25 |
| 330849 | 1844140 | 4/19/2017 | 0.25 | 0.5 |
| 331108 | 1844140 | 4/20/2017 | 0.25 | 0.75 |
| 331641 | 1844140 | 4/24/2017 | 0.25 | 1 |
| 331736 | 1844140 | 4/25/2017 | 0.25 | 1.25 |
| 333761 | 1844140 | 5/5/2017 | 0.25 | 1.5 |
| 336080 | 1844140 | 5/17/2017 | 1 | 2.5 |
| 349752 | 1844140 | 8/2/2017 | 0.25 | 3 | <--- this should be 2.75
| 350994 | 1844140 | 8/9/2017 | 1 | 4 |
| 351426 | 1844140 | 8/11/2017 | 0.25 | 4.25 |
| 352132 | 1844140 | 8/15/2017 | 0.5 | 4.75 |
| 354236 | 1844140 | 8/25/2017 | 0.25 | 5 |
| 355580 | 1844140 | 8/29/2017 | 0.25 | 5.25 |
| 355650 | 1844140 | 9/5/2017 | 0.25 | 5.5 |
+--------+------------+-----------+-------+-------+
这个想法是它将遍历所有事件并总结在当前日期的一年内发生的所有先前事件。
使用我当前的代码,计算中存在总命中3的问题。
我想使用SQL Over语句,因为它正确计算但我不确定如何在滚动的1年时间范围内使用它。这是我用过的:
SELECT lngTIMEID as Id,
lngEMPID as EmployeeId,
dtmdateapp AS [Date],
stroccur AS [Value],
sum(strOCCUR) OVER(ORDER BY dtmdateapp ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) as [Total]
FROM [Attendance].[dbo].[timeuse] d
WHERE lngEMPID = 1844140 AND absence = 'Unscheduled' AND lngRID IN (1,2,3,4,5,7,8,9,10,11,12,13,14)
并给我正确的结果,因为这些数据都发生在同一年,但它不适用于滚动的1年时间框架:
+--------+------------+-----------+-------+-------+
| Id | EmployeeId | Date | Value | Total |
+--------+------------+-----------+-------+-------+
| 330435 | 1844140 | 4/17/2017 | 0.25 | 0.25 |
| 330849 | 1844140 | 4/19/2017 | 0.25 | 0.5 |
| 331108 | 1844140 | 4/20/2017 | 0.25 | 0.75 |
| 331641 | 1844140 | 4/24/2017 | 0.25 | 1 |
| 331736 | 1844140 | 4/25/2017 | 0.25 | 1.25 |
| 333761 | 1844140 | 5/5/2017 | 0.25 | 1.5 |
| 336080 | 1844140 | 5/17/2017 | 1 | 2.5 |
| 349752 | 1844140 | 8/2/2017 | 0.25 | 2.75 |
| 350994 | 1844140 | 8/9/2017 | 1 | 3.75 |
| 351426 | 1844140 | 8/11/2017 | 0.25 | 4 |
| 352132 | 1844140 | 8/15/2017 | 0.5 | 4.5 |
| 354236 | 1844140 | 8/25/2017 | 0.25 | 4.75 |
| 355580 | 1844140 | 8/29/2017 | 0.25 | 5 |
| 355650 | 1844140 | 9/5/2017 | 0.25 | 5.25 |
+--------+------------+-----------+-------+-------+
如何将1年滚动时间帧添加到MS SQL的Over语句中?
答案 0 :(得分:1)
正如詹姆斯所说,你的原始子查询应该有效。我在我的环境中测试了它并返回了正确的结果。
WITH tu AS (
SELECT *
FROM [Attendance].[dbo].[timeuse] d
WHERE lngEMPID = 1844140 AND absence = 'Unscheduled' AND lngRID IN (1,2,3,4,5,7,8,9,10,11,12,13,14)
)
SELECT lngTIMEID as Id,
lngEMPID as EmployeeId,
dtmdateapp AS [Date],
stroccur AS [Value],
(SELECT Sum(stroccur)
FROM tu a
WHERE a.dtmdateapp between DateAdd(d, -366, d.dtmdateapp)
AND d.dtmDATEAPP
AND a.lngempid = d.lngempid) AS Total
FROM tu d