I guess you can change a little the query to your needs.
The problem is pretty much the same...
I've used c3 column in equality, but if this column
has repeated values, just choose any column or combination of columns which
is unique.
Best,
Oliveiros
SELECT a.c1,a.c2,b.c3,b.c4,a.c5,b.c6
FROM
(
SE
Just add two conditions on the previous query
A particularity of this approach is that the non-null record will always
appear with the first child in alphabetical order.
But, according to what you explain, I guess it is OK, and if it doesn't it
is easily changed. :-)
Also I've changed the first c
hi,
i don't want any sort.. just like this example
*num father child age col5
*1 joe bruce14 8
lei 10
mike 5
2manuel child135 16
child233
child3 30
this is what
s
SELECT a.num,a.father,b.child,b.age
FROM
(
SELECT num,father, MAX(age)as maximo
FROM t1
GROUP BY num,father) a
RIGHT JOIN t1 b
ON b.age = a.maximo
- Original Message -
From: Zied Kharrat
To: pgsql-sql@postgresql.org
Sent: Thursday, October 23, 2008 9:14 AM
Subject: [SQL] Po
Hi Everybody..
Let's present my problem:
I have a table named *t1* and i will insert differents values like this :
insert into t1 (num,father,child,age) values ('1','joe','bruce','14',);
insert into t1 (num,father,child,age) values ('1','joe','lei','10',);
insert into t1 (num,father,child,age) v
