Question

我是新手。我有一个存储User_id的table1和另一个存储data_id的table2 将从table2中随机选择Data_id，并且所有user_id都应具有所有随机data_id。

                          table1                table2    
                          user1                 data1
                          user2                 data2
                          user3                 data3
                          user4                 data4
                          user5                 data5

我想从table2获取3个随机数据，并将这些数据与user_id #s一起写入table3 它应该如下所示

                                  table3
                              user1   data2                                                     
                              user1   data3
                              user1   data4  
                              user2   data2                                                     
                              user2   data3
                              user2   data4     
                              user3   data2                                                     
                              user3   data3
                              user3   data4  
                              user4   data2                                                     
                              user4   data3
                              user4   data4  
                              user5   data2                                                     
                              user5   data3
                              user5   data4

Answer 1

假设您希望每个用户使用相同的3个随机data_id值，则一个方法使用CROSS APPLY：

CREATE TABLE dbo.table1 ( user_id varchar(10) NOT NULL )
INSERT  INTO dbo.table1
VALUES  ( 'user1' )
    ,   ( 'user2' )
    ,   ( 'user3' )
    ,   ( 'user4' )
    ,   ( 'user5' );
CREATE TABLE dbo.table2 ( data_id varchar(10) NOT NULL );
INSERT  INTO dbo.table2
VALUES  ( 'data1' )
    ,   ( 'data2' )
    ,   ( 'data3' )
    ,   ( 'data4' )
    ,   ( 'data5' );

SELECT  user_id
      , data_id
FROM    dbo.table1
        CROSS APPLY ( SELECT TOP ( 3 )
                                data_id
                      FROM      dbo.table2
                      ORDER BY  NEWID()
                    ) AS table2
ORDER BY user_id
      , data_id;

Answer 2

基于@Dan Guzman回答我认为它已经参考并重新修改了答案。如果它适合您，请检查并告诉我

declare @table1 TABLE  ( user_id varchar(10) NOT NULL )
INSERT  INTO @table1
VALUES  ( 'user1' )
    ,   ( 'user2' )
    ,   ( 'user3' )
    ,   ( 'user4' )
    ,   ( 'user5' );
declare @table2 TABLE ( data_id varchar(10) NOT NULL );
INSERT  INTO @table2
VALUES  ( 'data1' )
    ,   ( 'data2' )
    ,   ( 'data3' )
    ,   ( 'data4' )
    ,   ( 'data5' );
;with cte as (
SELECT  user_id
      , data_id
FROM    @table1
        , ( SELECT TOP ( 4 )
                                data_id
                      FROM      @table2
                      ORDER BY data_id
                    ) AS table2
) 
  Select A.user_id,A.data_id from (    
 select   c.user_id,
          c.data_id,
          ROW_NUMBER()OVER(PARTITION BY c.user_id ORDER BY c.user_id,c.data_id) RN
 from cte C  

 )A
     where A.Rn <> 1

数据库随机到多对多

2 个答案: