在这上面看不到树木的木材,我确信它很简单。 我试图在连接表中返回相关记录的最大ID
表1
NiD Name
1 Peter
2 John
3 Arthur
表2
ID NiD Value
1 1 5
2 2 10
3 3 10
4 1 20
5 2 15
最大结果
NiD ID Value
1 4 20
2 5 15
3 3 10
答案 0 :(得分:0)
您可以使用row_number()
:
select NiD, ID, Value
from (select t2.*,
row_number() over (partition by NiD order by ID desc) as seqnum
from table2 t2
) t2
where seqnum = 1;
如上所述,您不需要table1
,因为table2
包含所有ID。
答案 1 :(得分:0)
CREATE TABLE Names
(
NID INT,
[Name] VARCHAR(MAX)
)
CREATE TABLE Results
(
ID INT,
NID INT,
VALUE INT
)
INSERT INTO Names VALUES (1,'Peter'),(2,'John'),(3,'Arthur')
INSERT INTO Results VALUES (1,1,5),(2,2,10),(3,3,10),(4,1,20),(5,2,15)
SELECT a.NID,
r.ID,
a.MaxVal
FROM (
SELECT NID,
MAX(VALUE) as MaxVal
FROM Results r
GROUP BY NID
) a
JOIN Results r
ON a.NID = r.NID AND a.MaxVal = r.VALUE
ORDER BY NID
答案 2 :(得分:0)
我就是这样做的,当Table2没有Table1记录的相应条目时,我认为ID
和Value
将是NULL
:
SELECT NiD, ID, [Value]
FROM Table1
OUTER APPLY (
SELECT TOP 1 ID, [Value]
FROM Table2
WHERE Table1.NiD = Table2.NiD
ORDER BY [Value] DESC
) AS Top_Table2
答案 3 :(得分:0)
这是我在类似情况下使用的,性能很好,前提是数据集不是太大(在1M行以下)。
SELECT
table1.nid
,table2.id
,table2.value
FROM table1
INNER JOIN table2 ON table1.nid = table2.nid
WHERE table2.value = (
SELECT MAX(value)
FROM table2
WHERE nid = table1.nid)
ORDER BY 1