我正在寻找一个查询，其中列出了24小时营业时间中给定时段内出现的人数。这是例如表
Customer_ID ENTRYDATETIME EXITDATETIME
-----------------------------------------------------------------
1001 2017-01-01 09:06:35.477 2017-01-01 12:06:35.477
1001 2017-01-05 11:06:35.477 2017-01-05 17:05:00.000
1002 2017-01-01 14:06:35.477 2017-01-01 19:05:00.000
1003 2017-01-01 14:06:35.477 2017-01-01 20:05:00.000
1067 2017-01-01 14:06:35.477 2017-01-01 16:05:00.000
1067 2017-01-02 14:06:35.477 2017-01-01 15:05:00.000
1067 2017-01-03 14:06:35.477 2017-01-01 16:05:00.000
1091 2017-01-10 14:06:35.477 2017-01-01 18:05:00.000
所以在这里你看到客户1001出现在1stJAN上午9点，上午10点，上午11点，下午12点
同样，客户＃1002在1stJAN上午11点，12,13,14,1500出现。
在本月底，我希望看到商店中每小时的平均客户数量。甚至平均每月等等。
有人可以帮我逻辑一下。非常感谢您的时间和精力。
答案 0 :(得分：2)
你可以使用它。
DECLARE @Temp TABLE (Customer_ID INT, ENTRYDATETIME DATETIME, EXITDATETIME DATETIME)
INSERT INTO @Temp VALUES
(1001,'2017-01-01 09:06:35.477','2017-01-01 12:06:35.477'),
(1001,'2017-01-05 11:06:35.477','2017-01-05 17:05:00.000'),
(1002,'2017-01-01 14:06:35.477','2017-01-01 19:05:00.000'),
(1003,'2017-01-01 14:06:35.477','2017-01-01 20:05:00.000'),
(1067,'2017-01-01 14:06:35.477','2017-01-01 16:05:00.000'),
(1067,'2017-01-02 14:06:35.477','2017-01-02 15:05:00.000'),
(1067,'2017-01-03 14:06:35.477','2017-01-03 16:05:00.000'),
(1091,'2017-01-10 14:06:35.477','2017-01-10 18:05:00.000')
;WITH CTE as
(
select Customer_ID, CAST(ENTRYDATETIME AS DATE) ENTRYDATE, DATEPART(HOUR, ENTRYDATETIME) ENTRYHOUR from @Temp
UNION all
select CTE.Customer_ID, CAST(ENTRYDATETIME AS DATE) DT, ENTRYHOUR = ENTRYHOUR + 1 from @Temp T INNER JOIN CTE ON T.Customer_ID = CTE.Customer_ID
AND ENTRYDATE = CAST(ENTRYDATETIME AS DATE)
AND (ENTRYHOUR + 1) <= DATEPART(HOUR, EXITDATETIME)
)
SELECT * FROM CTE
order by Customer_ID, ENTRYDATE, ENTRYHOUR
第二种方法：我使用了带时间表的交叉连接，而不是递归CTE。可能比第一种方法工作得更快。
;WITH HourTable AS
(
SELECT * FROM (VALUES (0),(1),(2),(3),(4),(5),(6),(7),(8),(9),(10),(11),(12),(13),(14),(15),(16),(17),(18),(19),(20),(21),(22),(23)) Hhr (HourIndex)
)
SELECT
T.Customer_ID,
CAST(T.ENTRYDATETIME AS DATE) ENTRYDATE, DATEPART(HOUR, DATEADD(HOUR, H.HourIndex, T.ENTRYDATETIME)) ENTRYHOUR
FROM @Temp T
INNER JOIN HourTable H ON DATEDIFF( HOUR, T.ENTRYDATETIME, T.EXITDATETIME ) >= H.HourIndex
ORDER BY
Customer_ID,
ENTRYDATE,
ENTRYHOUR
结果：
Customer_ID ENTRYDATE ENTRYHOUR
----------- ---------- -----------
1001 2017-01-01 9
1001 2017-01-01 10
1001 2017-01-01 11
1001 2017-01-01 12
1001 2017-01-05 11
1001 2017-01-05 12
1001 2017-01-05 13
1001 2017-01-05 14
1001 2017-01-05 15
1001 2017-01-05 16
1001 2017-01-05 17
1002 2017-01-01 14
1002 2017-01-01 15
1002 2017-01-01 16
1002 2017-01-01 17
1002 2017-01-01 18
1002 2017-01-01 19
1003 2017-01-01 14
1003 2017-01-01 15
1003 2017-01-01 16
1003 2017-01-01 17
1003 2017-01-01 18
1003 2017-01-01 19
1003 2017-01-01 20
1067 2017-01-01 14
1067 2017-01-01 15
1067 2017-01-01 16
1067 2017-01-02 14
1067 2017-01-02 15
1067 2017-01-03 14
1067 2017-01-03 15
1067 2017-01-03 16
1091 2017-01-10 14
1091 2017-01-10 15
1091 2017-01-10 16
1091 2017-01-10 17
1091 2017-01-10 18
这就是按小时计算的。
;WITH CTE as
(
select Customer_ID, CAST(ENTRYDATETIME AS DATE) ENTRYDATE, DATEPART(HOUR, ENTRYDATETIME) ENTRYHOUR from @Temp
UNION all
select CTE.Customer_ID, CAST(ENTRYDATETIME AS DATE) DT, ENTRYHOUR = ENTRYHOUR + 1 from @Temp T INNER JOIN CTE ON T.Customer_ID = CTE.Customer_ID
AND ENTRYDATE = CAST(ENTRYDATETIME AS DATE)
AND (ENTRYHOUR + 1) <= DATEPART(HOUR, EXITDATETIME)
)
SELECT ENTRYHOUR, COUNT(*) AVARAGE FROM CTE
GROUP BY ENTRYHOUR
结果：
ENTRYHOUR AVARAGE
----------- -----------
9 1
10 1
11 2
12 2
13 1
14 7
15 7
16 6
17 4
18 3
19 2
20 1
答案 1 :(得分：0)
我建议您生成date / time表格（使用CTE或静态表格），然后在＆＃39; BETWEEN＆＃39;上加入该表格。 EntryDateTime和ExitDateTime。