Обсуждение: cartesian product

Hi,

I have a query regarding an SQL statement I'm trying to execute. I
have the following table:

sequence
-------------+
AK
AKCMK
CMKA

I execute the following statement (Cartesian product):

SELECT p1.sequence as sequence1, p2.sequence as sequence2
FROM potential_pairs p1, potential_pairs p2

which gives me:

sequence1 | sequence2
----------------+--------------AK           | AK AK           | AKCMK AK           | CMKA AKCMK    | AK AKCMK    |
AKCMKAKCMK    | CMKA CMKA      | AK CMKA      | AKCMK CMKA      | CMKA
 
(9 rows)

I want to eliminate duplicates and by duplicate I mean a tuple such as
{AK, CMKA} should be regarded as the same as {CMKA, AK}. So I would
like the following result:

sequence1 | sequence2
----------------+--------------AK           | AK AK           | AKCMK AK           | CMKA AKCMK    | AKCMK AKCMK    |
CMKACMKA      | CMKA
 

Any help would be appreciated.

- Salman

On Mon, 2007-02-19 at 10:58, Salman Tahir wrote:
> Hi,
> 
> I have a query regarding an SQL statement I'm trying to execute. I
> have the following table:
> 
> sequence
> -------------+
> AK
> AKCMK
> CMKA
> 
> I execute the following statement (Cartesian product):
> 
> SELECT p1.sequence as sequence1, p2.sequence as sequence2
> FROM potential_pairs p1, potential_pairs p2
> 
> which gives me:
> 
> sequence1 | sequence2
> ----------------+--------------
>  AK           | AK
>   AK           | AKCMK
>   AK           | CMKA
>   AKCMK    | AK
>   AKCMK    | AKCMK
>   AKCMK    | CMKA
>   CMKA      | AK
>   CMKA      | AKCMK
>   CMKA      | CMKA
> (9 rows)
> 
> I want to eliminate duplicates and by duplicate I mean a tuple such as
> {AK, CMKA} should be regarded as the same as {CMKA, AK}. So I would
> like the following result:
> 
> sequence1 | sequence2
> ----------------+--------------
>  AK           | AK
>   AK           | AKCMK
>   AK           | CMKA
>   AKCMK    | AKCMK
>   AKCMK    | CMKA
>   CMKA      | CMKA
> 


SELECT p1.sequence as sequence1, p2.sequence as sequence2 FROM
potential_pairs p1, join potential_pairs p2 on
(p1.sequence<>p2.sequence) where sequence1 > sequence2

On Mon, 2007-02-19 at 10:58, Salman Tahir wrote:
> Hi,
> 
> I have a query regarding an SQL statement I'm trying to execute. I
> have the following table:
> 
> sequence
> -------------+
> AK
> AKCMK
> CMKA
> 
> I execute the following statement (Cartesian product):
> 
> SELECT p1.sequence as sequence1, p2.sequence as sequence2
> FROM potential_pairs p1, potential_pairs p2
> 
> which gives me:
> 
> sequence1 | sequence2
> ----------------+--------------
>  AK           | AK
>   AK           | AKCMK
>   AK           | CMKA
>   AKCMK    | AK
>   AKCMK    | AKCMK
>   AKCMK    | CMKA
>   CMKA      | AK
>   CMKA      | AKCMK
>   CMKA      | CMKA
> (9 rows)
> 
> I want to eliminate duplicates and by duplicate I mean a tuple such as
> {AK, CMKA} should be regarded as the same as {CMKA, AK}. So I would
> like the following result:
> 
> sequence1 | sequence2
> ----------------+--------------
>  AK           | AK
>   AK           | AKCMK
>   AK           | CMKA
>   AKCMK    | AKCMK
>   AKCMK    | CMKA
>   CMKA      | CMKA

Oh wait, slightly different thing you meant.  OK, you'll want something
like:

select p1.sequence as sequence1, p2.sequence as sequence2 from 
potential_pairs p1, potential_pairs p2 where p1.sequence >= p2.sequence 

Hi, Salman

Maybe this isn't so much elegant, but works:

SELECT p1.sequence as sequence1, p2.sequence as sequence2
FROM potential_pairs p1, potential_pairs p2
WHERE p1.sequence <= p2.sequence

Hope this helps


Salman Tahir wrote:
> Hi,
>
> I have a query regarding an SQL statement I'm trying to execute. I
> have the following table:
>
> sequence
> -------------+
> AK
> AKCMK
> CMKA
>
> I execute the following statement (Cartesian product):
>
> SELECT p1.sequence as sequence1, p2.sequence as sequence2
> FROM potential_pairs p1, potential_pairs p2
>
> which gives me:
>
> sequence1 | sequence2
> ----------------+--------------
> AK           | AK
>  AK           | AKCMK
>  AK           | CMKA
>  AKCMK    | AK
>  AKCMK    | AKCMK
>  AKCMK    | CMKA
>  CMKA      | AK
>  CMKA      | AKCMK
>  CMKA      | CMKA
> (9 rows)
>
> I want to eliminate duplicates and by duplicate I mean a tuple such as
> {AK, CMKA} should be regarded as the same as {CMKA, AK}. So I would
> like the following result:
>
> sequence1 | sequence2
> ----------------+--------------
> AK           | AK
>  AK           | AKCMK
>  AK           | CMKA
>  AKCMK    | AKCMK
>  AKCMK    | CMKA
>  CMKA      | CMKA
>
> Any help would be appreciated.
>
> - Salman
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
>       choose an index scan if your joining column's datatypes do not
>       match
>
>

-- 
Luiz K. Matsumura
Plan IT Tecnologia Informática Ltda.