Self join on a table with condition to extract records

Question

Hello PL/SQL Gurus/experts,
I am using Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production version
I have following table -
drop table t1;
create table t1(stud_NM,stud_id) as select
'Vikas',1 from dual union all select
'Andy',2 from dual union all select
'Chi',3 from dual union all select
'Derek',4 from dual union all select
'Kristien',5 from dual union all select
'Tom',6 from dual union all select 
'Messi',7 from dual union all select
'Ketty',8 from dual;
drop table t2;
create table t2(stud_id,Class, CDate) as select
1,'BA','20120101' from dual union all select
1,'MA','20120101' from dual union all select
2,'BSc','20120531' from dual union all select
3,'MCA','20120606' from dual union all select
4,'MCA','20120701' from dual union all select
4,'Btech','20120715' from dual union all select
5,'MTech','20120730' from dual union all select
6,'MSc','20120801' from dual union all select
6,'MSc','20120802' from dual union all select   
7,'MA','20120815' from dual union all select
8,'MA','20120815' from dual;
I have to get the output based on condition -
Extract Stud_Name, Count of records which has a condition as that of the one that if combination of Stud_Name & CDate is same then count this combination once only else if Stud_Name & CDate are different then count them as different records -
Currently i am using the following query to get the output -
select t1.stud_NM, count(distinct CASE WHEN orig.CDate<>ali.CDate then t1.stud_NM else
CASE WHEN orig.CDate=ali.CDate then '1' else null
end
end) Count
from t1,
t2 orig,
t2 ali
where t1.stud_id=orig.stud_id
and t1.stud_id=ali.stud_id
group by t1.stud_NM
order by t1.stud_NM
which is giving the following output, which looks fine as well -
STUD_NM       COUNT
Andy              1
Chi               1
Derek             2
Ketty             1
Kristien          1
Messi             1
Tom               2
Vikas             1
8 rows selected.
So want to confirm if i am doing right or it is just because of the data it is giving me the right result ?
Kindly help me thanks in advance for your time and efforts.

Answer 1

Hi,
You can simplify that, too.  Since you're GROUPing BY stud_nm, concatenating COUNT (DISTINCT stud_nm || cdate) is just the same as COUNT (DISTINCT cdate), assuming cdate is not NULL.
Also, that assumes stud_nm is unique.  (If it's not, the desired output doesn't make sense, but that's a different matter.)  I'd do it this way:
SELECT    t1.stud_nm
,         t1.stud_id    -- If wanted
,         COUNT (DISTINCT t2.cdate)  AS cnt
FROM             t1
LEFT OUTER JOIN  t2  ON  t1.stud_id  = t2.stud_id
GROUP BY  t1.stud_nm, t1.stud_id
ORDER BY  t1.stud_nm, t1.stud_id
OP: the query you posted will get the wrong results with other data.  For example, if we add this to the sample data in t2:
7,'WX','20120816' from dual union all select
7,'YZ','20120817' from dual union all select
Karthick's solution (and my variant of it) correctly show a count of 3 for stud_id=7 (Messi), but your query has 2.