我很好奇是否可以使用窗口函数按日期范围对数据进行分区？我有表TerritoryRevenue：
TerritoryID name Revenue orderdate RankByRevenue
----------- -------------------------------------------------- --------------------- ----------------------- --------------------
4 Southwest 1027923,9413 2006-09-01 00:00:00.000 1
4 Southwest 892043,959 2006-08-01 00:00:00.000 2
4 Southwest 796734,8925 2007-05-01 00:00:00.000 3
4 Southwest 782962,2628 2006-11-01 00:00:00.000 4
4 Southwest 775273,9148 2007-09-01 00:00:00.000 5
4 Southwest 760638,3393 2007-07-01 00:00:00.000 6
1 Northwest 733397,8075 2007-08-01 00:00:00.000 7
6 Canada 713068,2334 2007-12-01 00:00:00.000 8
7 France 707659,8229 2007-08-01 00:00:00.000 9
6 Canada 677073,7101 2007-09-01 00:00:00.000 10
4 Southwest 676897,9351 2006-12-01 00:00:00.000 11
4 Southwest 672075,0706 2007-08-01 00:00:00.000 12
6 Canada 660610,2897 2006-08-01 00:00:00.000 13
4 Southwest 644851,3869 2007-02-01 00:00:00.000 14
7 France 640657,4244 2007-11-01 00:00:00.000 15
6 Canada 638115,2008 2006-07-01 00:00:00.000 16
4 Southwest 630143,7952 2007-06-01 00:00:00.000 17
4 Southwest 623321,4611 2007-10-01 00:00:00.000 18
7 France 616132,3947 2008-05-01 00:00:00.000 19
1 Northwest 606621,8131 2007-09-01 00:00:00.000 20
4 Southwest 598632,0709 2008-06-01 00:00:00.000 21
1 Northwest 597304,4287 2007-11-01 00:00:00.000 22
4 Southwest 591577,4069 2008-04-01 00:00:00.000 23
4 Southwest 589925,8656 2008-05-01 00:00:00.000 24
6 Canada 589015,6507 2008-06-01 00:00:00.000 25
4 Southwest 588013,9785 2007-12-01 00:00:00.000 26
4 Southwest 578002,5452 2007-11-01 00:00:00.000 27
6 Canada 572102,5723 2007-08-01 00:00:00.000 28
1 Northwest 567438,8724 2008-05-01 00:00:00.000 29
6 Canada 564052,7259 2006-09-01 00:00:00.000 30
1 Northwest 555031,8462 2006-08-01 00:00:00.000 31
1 Northwest 524780,7777 2008-06-01 00:00:00.000 32
1 Northwest 521845,3013 2007-12-01 00:00:00.000 33
4 Southwest 517015,445 2008-02-01 00:00:00.000 34
1 Northwest 516421,3597 2006-11-01 00:00:00.000 35
6 Canada 510381,3556 2006-11-01 00:00:00.000 36
6 Canada 502797,2246 2007-07-01 00:00:00.000 37
10 United Kingdom 496271,7803 2007-09-01 00:00:00.000 38
4 Southwest 496217,5327 2006-07-01 00:00:00.000 39
6 Canada 489254,911 2006-12-01 00:00:00.000 40
1 Northwest 483485,0034 2007-05-01 00:00:00.000 41
7 France 477976,7538 2008-02-01 00:00:00.000 42
4 Southwest 474840,9012 2008-01-01 00:00:00.000 43
6 Canada 469752,6114 2007-05-01 00:00:00.000 44
6 Canada 467334,1973 2008-03-01 00:00:00.000 45
4 Southwest 466123,0048 2008-03-01 00:00:00.000 46
1 Northwest 456366,7061 2008-02-01 00:00:00.000 47
6 Canada 443495,2247 2007-06-01 00:00:00.000 48
1 Northwest 441917,9528 2008-03-01 00:00:00.000 49
4 Southwest 438879,4098 2007-03-01 00:00:00.000 50
(...)
如您所见，我已经按Revenue对我的输出进行了排名，但是没有 partition by 子句。我想在两个时间间隔内分别按Revenue进行排名，比如说从2006年7月到2007年6月以及从2007年7月到2008年6月。
答案 0 :(得分：1)
尝试下面的查询，您可以使用case when语句生成分组列：
select TerritoryID,
name,
Revenue,
orderdate,
row_number() over (partition by [06-08] + [06-07] order by Revenue) RankByRevenue
from (
select TerritoryID,
name,
Revenue,
orderdate,
case when orderdate between '2006-07-01' and '2007-01-01' then 1 else 0 end [06-07],
case when orderdate between '2007-07-01' and '2008-01-01' then -1 else 0 end [07-08]
from MyTable
) a