普适数据

Question

I have a table with in there following data

 ID|LabelID|Value 1 |1 |3 1 |2 |1 1 |3 |15 2 |1 |5 2 |2 |7 2 |3 |5 

I want to get the following as result within a pervasive database

 ID|Label1|Label2|Label3 1 |3 |1 |15 2 |5 |7 |5 

Anyone an idea? I did try some things and the best result I can get is the following:

 ID|Label1|Label2|Label3 1 |3 | | 1 | |1 | 1 | | |15 2 |5 | | 2 | |7 | 2 | | |5 

Answer 1

Have fun... :-)

Should work in most versions of SQL. There are vendor specific PIVOT statements which are also an option.

This was the way to do it prior to PIVOT clause.

drop table #test;
create table #test (ID int, LabelID int, Value int);

insert into #test values (1, 1, 3)
,(1, 2, 1)
,(1, 3, 15)
,(2, 1, 5)
,(2, 2, 7)
,(2, 3, 5);

select ID
      ,sum(case when LabelID = 1 then Value else null end) as Label1
      ,sum(case when LabelID = 2 then Value else null end) as Label2
      ,sum(case when LabelID = 3 then Value else null end) as Label3
  from #test
group by ID;

Answer 2

Here's a query that worked for me.

create table psqlPivot (id integer, labelid integer, val integer);
insert into psqlPivot values (1, 1,3);
insert into psqlPivot values (1, 2,1);
insert into psqlPivot values (1, 3,15);
insert into psqlPivot values (2, 1,5);
insert into psqlPivot values (2, 2,7);
insert into psqlPivot values (2, 3,5);

select distinct v.id, (select a.val as label1 from psqlPivot a where a.labelid = 1 and a.id = v.id)
,(select b.val as label2 from psqlPivot b where b.labelid = 2 and b.id = v.id)
,(select c.val as label3 from psqlPivot c where c.labelid = 3 and c.id = v.id)
from psqlpivot v