过滤联接

Question

我有两个桌子。

表1

  -----RID------URID-----Status-----
       1         1        0
       1         2        1
       1         3        3
       2         4        0

表2

 -----RID----Success------Errors---
       1        2            1
       2        1            0

我想为表1中的每个名称计算表2中的所有错误（状态> 1），所有成功（状态<2）。像这样：

 select distinct T1.[Name]
 ,count(T2.URID) as Error
 ,count(T3.URID) as Success
 FROM [dbo].[Table 1] T1
 left join [dbo].[Table 2] T2
 on T2.RID = T1.RID and T2.Status > 1
 left join [dbo].[Table 2] T3
 on T3.RID = T1.RID and T3.Status < 2
 group by RID   

这是我要使用的查询，但不起作用...认为在联接中已经过滤会很明智。...我在这里做错了什么？

dropApples(){
      let apple;
      var random = 2;
      if(this.state.apples){
        this.setState(prevState => ({
            apples: prevState.apples.map(apple => {
                if (apple.id == random) return {...apple, status: 'falling'}
                return apple
            })
        })) 
    }
} 

Answer 1

您也可以通过以下方式实现这一目标，

 SELECT * INTO #T FROM (SELECT CASE WHEN T.[STATUS]>1 THEN COUNT(T.[STATUS]) ELSE 0 END [ERROR],
 CASE WHEN T.[STATUS]<2 THEN COUNT(T.[STATUS]) ELSE 0 END [SUCCESS],
 T2.NAME FROM TEST1 T2 INNER JOIN TEST2 T ON T.RID=T2.RID
 GROUP BY T.[STATUS],T2.[NAME]) T


 SELECT NAME,SUM([SUCCESS]) [SUCCESS],SUM(ERROR)[ERROR] FROM #T GROUP BY NAME

Answer 2

您可以进行聚合：

select  t1.rid, t1.name,
        sum(case when status > 1 then 1 else 0 end) as error, 
        sum(case when status < 2 then 1 else 0 end) as success 
from table1 t1 inner join
     table2 t2
     on t2.rid = t1.rid
group by t1.rid, t1.name

但是，如果您有DISTINCT子句，则此处不需要GROUP BY。

Answer 3

使用条件聚合：

select T1.[Name],
       sum(case when t2.status > 1 then 1 else 0 end) as errors,
       sum(case when t2.status < 2 then 1 else 0 end) as success
from [dbo].[Table 1] T1 left join
     [dbo].[Table 2] T2
     on T2.RID = T1.RID
group by T1.RID, T1.[Name];

通过用于定义每一行T1.Name的列进行聚合也很重要。我不确定T1.RID中是否需要group by，但是您的版本中是否有它。

Answer 4

这将起作用：

create table ns_999(col1 number,col2 number,col3 number);

 insert into ns_999 values( 1,1,0);
 insert into ns_999 values(1,2 ,1);
 insert into ns_999 values(1,3,3);
  insert into ns_999 values(2,4,0);
  SELECT * FROM ns_999;

  select col1,
       sum(case when col3>1 then 1 else 0 end ) as errors, 
       sum(case when col3<2 then 1 else 0 end) as sucess
from ns_999 
group by col1;

示例输出：

 RID FALIURE SUCCESS 
    1      1       2
    2      0       1