Question

我的表格数据如下： UserData表：

|ID| Name  | TeamName |
| 1| Peter | Alpha    |
| 1| Peter | Beta     |
| 1| Peter | Gamma    |
| 2| Mary  | Gamma    |
| 2| Mary  | Omega    |
| 3| John  | Kappa    |
| 3| John  | Delta    |

Name和TeamName的组合始终是唯一的。我需要每个唯一的ID和名称获得前1个TeamName和团队关系的数量，如下所示：

table #FinalTable
|ID| Name  | TeamName | NumberOfRelations |
| 1| Peter | Alpha    | 3                 |
| 2| Mary  | Gamma    | 2                 |
| 3| John  | Kappa    | 2                 |

问题 - 有没有办法在一个查询中执行此操作，或者我是否必须使用临时表来选择前1个团队并计算关系数量，然后选择数据indo单独的最终表？我试过这样的事情：

;WITH cte AS
(
   SELECT *, ROW_NUMBER() OVER (PARTITION BY ID ORDER BY TeamName Asc) AS rn
   FROM UserData
)
SELECT * into #tempTable1
FROM cte
WHERE rn = 1

和此：

insert into #tempTable2 (ID, Name, NumberOfRelations)
select ID, Name, count(*) as NumberOfRelations
from UserData
group by ID, Name

...然后从两个临时表中选择数据。我想知道是否有更简单的方法。

Answer 1

对于SQLserver：

你没有订购，所以我在下面选择一个......

select top 1 with ties id,playen,count(id) over (partition by id,playen) as countt
,temaname
from #temp t1
order by row_number() over (partition by id,playen  order by id,playen,temaname)

<强>输出：

id  playen  countt  temaname
1   Peter   3       Alpha
2   Mary    2      Gamma
3   John    2      Delta

Answer 2

假设这是SQL Server，请尝试：

    Select t.ID, t.Name, team.TeamName, count(t.TeamName) countt
from @temp t join
(Select id, TeamName, Row_Number() over (Partition By ID Order By TeamName asc) as rn
from @temp) team on (team.ID = t.ID and team.rn=1)
Group by t.ID, t.Name, team.TeamName

Answer 3

SQL表代表无序集。除非列指定排序，否则没有第一个团队名称。你似乎没有这样的专栏。

如果您有这样的专栏：

WITH cte AS (
      SELECT ud.*,
             ROW_NUMBER() OVER (PARTITION BY ID ORDER BY ??) as seqnum,
             COUNT(*) OVER (PARTITION BY ID) as cnt
      FROM UserData ud
     )
SELECT cte.*
FROM cte
WHERE seqnum = 1;

请注意??。这是指定获取团队名称的顺序。根据数据库的不同，您可以使用NULL或(SELECT NULL)获取任意团队名称。

选择具有聚合函数的前1行

3 个答案: