我为我的表创建了一个默认值为0的列。我想用正确的序列号更新该表。我可以用Row_Number吗?
我的表是:
PersonID | Code | Sequence Number
---------+--------+------------
10 | D112 | 0
10 | D112 | 0
10 | D112 | 0
10 | E110 | 0
10 | E110 | 0
10 | E110 | 0
10 | D112 | 0
10 | D112 | 0
10 | D112 | 0
10 | E110 | 0
10 | E110 | 0
10 | E110 | 0
我希望我的桌子像这样:
PersonID | Code | Sequence Number
---------+--------+------------
10 | D112 | 1
10 | D112 | 1
10 | D112 | 1
10 | E110 | 2
10 | E110 | 2
10 | E110 | 2
11 | M490 | 1
11 | M490 | 1
11 | M490 | 1
11 | N550 | 2
11 | N550 | 2
11 | N550 | 2
这是我的代码,但不确定这是否正确。
WITH CTE AS (
SELECT
t.Sequence Number,
ROW_NUMBER() OVER (PARTITION BY t.PersonID, t.Code ORDER BY t.PersonID) AS RN
FROM Table AS t )
UPDATE CTE
SET Sequence Number = RN
答案 0 :(得分:3)
DECLARE @Table AS TABLE (PersonId INt, Code CHAR(4), SequenceNumber INT)
INSERT INTO @Table VALUES (10,'D112',0),(10,'D112',0),(10,'D112',0),(10,'E110',0)
,(10,'E110',0),(10,'E110',0),(10,'D112',0),(10,'D112',0),(10,'D112',0),(10,'E110',0)
,(10,'E110',0),(10,'E110',0)
;WITH CTE AS (
SELECT
PersonId
,Code
,SequenceNumber
,DENSE_RANK() OVER (PARTITION BY t.PersonID ORDER BY t.Code ) AS RN
FROM
@Table AS t)
UPDATE cte
SET SequenceNumber = RN
SELECT *
FROm
@Table
您非常接近,只需要将您的PARTITION BY
和ORDER BY
略有不同,然后使用DENSE_RANK()
代替ROW_NUMBER()
来处理这些关系。