生成 DENSE_RANK() 而不使用 order by 子句 - SQL 服务器

Question

我是 SQL Server 的新手，任何人都可以指导我的查询，这对我很有帮助。

我需要像这个例子一样生成排名：

Report No               Rank 
----------------------------
18  3229                 1
89-0119-87               2
180261292                3
180261292                3
18  3107                 4
18A 7     A06            5
T18002415                6
T18002415                6
T18002415                6
0000000018-519           7
0000000018-519           7
0000000018-519           7
0000000018-519           7
0000000018-655           8
000000018-1002           9
000000018-1002           9
000000018-1002           9
000000018-5712           10

如果我使用

生成排名

DENSE_RANK() OVER(ORDER BY ReportNo DESC)

那么我没有得到所需的输出。你能指导我吗？我怎样才能像上面的例子那样生成排名？

Answer 1

我看不出报告编号排序的逻辑。因为它是自定义订单，所以创建 anotherTable （临时 table 或 table 变量）包含所需的订单 Report No ， Rank.

Report No           Rank 
18  3229                 1
89-0119-87               2
180261292                3
18  3107                 4
18A 7     A06            5
T18002415                6
0000000018-519           7
0000000018-655           8
000000018-1002           9
000000018-5712           10

然后

SELECT thisTable.ReportNo, anotherTable.[Rank]
FROM thisTable
INNER JOIN anotherTable ON thisTable.ReportNo = anotherTable.ReportNo

Answer 2

试试这个：

DECLARE @Tab TABLE(Report_No VARCHAR(100))

INSERT INTO @Tab VALUES('18  3229')
INSERT INTO @Tab VALUES('89-0119-87')
INSERT INTO @Tab VALUES('180261292')
INSERT INTO @Tab VALUES('180261292')
INSERT INTO @Tab VALUES('18  3107')
INSERT INTO @Tab VALUES('18A 7     A06')
INSERT INTO @Tab VALUES('T18002415')
INSERT INTO @Tab VALUES('T18002415')
INSERT INTO @Tab VALUES('T18002415')
INSERT INTO @Tab VALUES('0000000018-519')
INSERT INTO @Tab VALUES('0000000018-519')
INSERT INTO @Tab VALUES('0000000018-519')
INSERT INTO @Tab VALUES('0000000018-519')
INSERT INTO @Tab VALUES('0000000018-655')
INSERT INTO @Tab VALUES('000000018-1002')
INSERT INTO @Tab VALUES('000000018-1002')
INSERT INTO @Tab VALUES('000000018-1002')
INSERT INTO @Tab VALUES('000000018-5712')

SELECT *
    ,ROW_NUMBER() OVER(ORDER BY (SELECT NULL)) RN
INTO #temp
FROM @Tab T1

SELECT T1.Report_No
    ,SUM(CASE WHEN T1.Report_No = T2.Report_No THEN 0 ELSE 1 END) OVER(ORDER BY T1.rn)[Rank]
FROM #temp T1
LEFT JOIN #temp T2 ON T1.rn = T2.rn+1

DROP TABLE #temp

输出：

Report_No       Rank
18  3229        1
89-0119-87      2
180261292       3
180261292       3
18  3107        4
18A 7     A06   5
T18002415       6
T18002415       6
T18002415       6
0000000018-519  7
0000000018-519  7
0000000018-519  7
0000000018-519  7
0000000018-655  8
000000018-1002  9
000000018-1002  9
000000018-1002  9
000000018-5712  10

Answer 3

使用分组依据：

WITH data as (
    SELECT '18  3229' col1 UNION ALL
    SELECT '89-0119-87' UNION ALL
    SELECT '180261292' UNION ALL
    SELECT '180261292' UNION ALL
    SELECT '18  3107' UNION ALL
    SELECT '18A 7     A06' UNION ALL
    SELECT 'T18002415' UNION ALL
    SELECT 'T18002415' UNION ALL
    SELECT 'T18002415' UNION ALL
    SELECT '0000000018-519' UNION ALL
    SELECT '0000000018-519' UNION ALL
    SELECT '0000000018-519' UNION ALL
    SELECT '0000000018-519' UNION ALL
    SELECT '0000000018-655' UNION ALL
    SELECT '000000018-1002' UNION ALL
    SELECT '000000018-1002' UNION ALL
    SELECT '000000018-1002' UNION ALL
    SELECT '000000018-5712'), 
data_with_rownumber as (
    SELECT *, row_number() over (order by (SELECT 1)) rownum from data
), grouped as (
    SELECT col1, min(rownum) rank from data_with_rownumber group by col1)
select d.col1, DENSE_RANK() over (order by rank) from data d inner join grouped gr on d.col1 = gr.col1

生成 DENSE_RANK() 而不使用 order by 子句 - SQL 服务器

Generate DENSE_RANK() without using order by clause - SQL Server

tsql

database

sql-server

window-functions