Denormalizing Data (Maybe A Pivot?)

Question

I have a simple table containing Student Numbers and corresponding Teacher Numbers, and I need to denormalize it for input to a legacy system.

For example, here\'s

Answer 1

The old school method is to use CASE expressions; since SQL Server 2005 you can use PIVOT.

CASE example:

SELECT t.studnumber,
       CASE WHEN t.teachernumber = 57315 THEN t.teachernumber ELSE NULL END AS TEACHER1,
       CASE WHEN t.teachernumber = 88115 THEN t.teachernumber ELSE NULL END AS TEACHER1

But dynamically placing a given teachernumber as teacher1/etc is an entirely different matter.

Answer 2

I second PIVOT.

Here are two interesting links with programmatic solutions to dynamic crosstab data in SQL.

http://www.simple-talk.com/sql/t-sql-programming/creating-cross-tab-queries-and-pivot-tables-in-sql/

http://www.simple-talk.com/sql/t-sql-programming/crosstab-pivot-table-workbench/

Answer 3

You can use pivot. You also need to "Rank" your teachers 1-6. See my comment on how you want to do this. For now:

Select StudNumber, TeacherNumber, TeacherRank
from (
   Select ST.StudNumber
       , ST.TeacherNumber
       , ROW_NUMBER() OVER (PARTITION BY ST.StudNumber 
                    ORDER BY ST.TeacherNumber) AS TeacherRank
   From StudentTeacher AS ST)
Where TeacherRank <=6

Then you can pivot on this statement. Here is a good explanation: Using Pivot and UnPivot