How to get get Unique Records based on multiple columns from a table

Question

Consider the following table:

primaryKey   id    activity   template  creator   created
1            1      3           5         x       2011-10-13
2                


        

Answer 1

I think this should work -

SELECT * 
FROM TABLE
WHERE 
 primaryKey in 
 (
   SELECT min(primarkyKey) from TABLE 
      group by id, activity, template
 )

Here, first distinct is obtain on required columns in the inner query by doing group by. Then the min of primary key of each distinct record is used to get all the columns from the outer query.

Answer 2

This is for MS SQL Server.

Updated, as i made a little mistake!

SELECT DISTINCT 
        ROW_NUMBER() OVER (ORDER BY 
                                id    
                            ,   activity   
                            ,   template  
                            ,   creator  
                            ,   created ) PrimaryKey

    ,   id    
    ,   activity   
    ,   template  
    ,   creator  
    ,   created 
    FROM 
    [TABLE_NAME]
    GROUP BY 

        id    
    ,   activity   
    ,   template  
    ,   creator  
    ,   created 

Answer 3

SELECT primarykey, id, activity, template, creator, created FROM (
    SELECT *, row_number() OVER (partition BY id, activity, template ORDER BY created) as rn FROM table
) a 
WHERE rn = 1