Question

I have data in 1 table as

UserID CourseID TestID   Result
1         1       1        P
1         1       2        P
2         1       1        F
2         1       2        F

this is showing that 2 Users ( UserID =1 and 2) has taken Test ( TestID=1 and 2) in CourseID =1

Now that CourseID in total has 3 Tests ( say TestID=1,2,7) as

CourseID TestID
1          1
1          2
1          7

this means none of the user has taken Test ( TestID=7) show now i want to show data as below

UserID CourseID TestID   Result
1         1       1        P
1         1       2        P
1         1       7        null
2         1       1        F
2         1       2        F
2         1       7        null

I have been trying with left/right outer join with Group by, but could not get the desired result. How it can be achieved?

Answer 1

SQL Fiddle

MS SQL Server 2008 Schema Setup:

create table Result
(
  UserID int,
  CourseID int,
  TestID int,
  Result char(1)
)
insert into Result values
(1,         1,       1,        'P'),
(1,         1,       2,        'P'),
(2,         1,       1,        'F'),
(2,         1,       2,        'F')

create table Course
(
  CourseID int,
  TestID int
)

insert into Course values
(1, 1),
(1, 2),
(1, 7)

Query 1:

select U.UserID,
       C.CourseID,
       C.TestID,
       R.Result
from (
     select distinct UserID
     from Result
     ) as U
  cross apply Course as  C
  left outer join Result as R
    on R.CourseID = C.CourseID and
       R.TestID = C.TestID and
       R.UserID = U.UserID

Results:

| USERID | COURSEID | TESTID | RESULT |
---------------------------------------
|      1 |        1 |      1 |      P |
|      1 |        1 |      2 |      P |
|      1 |        1 |      7 | (null) |
|      2 |        1 |      1 |      F |
|      2 |        1 |      2 |      F |
|      2 |        1 |      7 | (null) |

Note: If you have a table Users you could replace the derived table

 (
 select distinct UserID
 from Result
 ) as U

with Users as U instead.