[SELECT] question

Hi all,
I've got a table structured as follow where the code field identify a particular parameter. Is it possible with a single query to retrieve more than just one parameter at a time? I mean can I get a recordset with the first field representing the date, the second field holding the values of parameter code [01], the third field the values of parameter code [02] and so forth??

Table structure
CREATE TABLE `test` (
`date` datetime NOT NULL ,
`code` varchar(3) NOT NULL ,
`value` float NOT NULL
) TYPE=MyISAM;

Sample data
+-------+--+---+
| date | code | value |
+-------+--+---+
| 2004-06-01 00:00:00 | 001 | 89 |
| 2004-06-01 00:00:00 | 002 | 1 |
| 2004-06-01 01:00:00 | 001 | 76 |
| 2004-06-01 01:00:00 | 002 | 5 |
| 2004-06-01 02:00:00 | 001 | 67 |
| 2004-06-01 02:00:00 | 002 | 7 |
| 2004-06-01 03:00:00 | 001 | 46 |
| 2004-06-01 03:00:00 | 002 | 4 |
| 2004-06-01 04:00:00 | 001 | 43 |
| 2004-06-01 04:00:00 | 002 | 3 |
+-------+--+---+

Recordset
+-------+----+----+
| date | value_cod1 | value_cod2 |
+-------+----+----+
| 2004-06-01 00:00:00 | 89 | 1 |
| 2004-06-01 01:00:00 | 76 | 5 |
| 2004-06-01 02:00:00 | 67 | 7 |
|..................................... ..............|
+-------+----+----+

Thank you very much!
Paolo SaudinTo select all of the columns in an arbitrary order, you can use:SELECT *
FROM testTo select the columns in a specific order, you can use:SELECT date, code, value
FROM test-PatP|||ah, the good old denormalization question

mysql doesn't support full outer join, but this is an equivalent --

select t1.date
, t1.value as value_cod1
, t2.value as value_cod2
from `test` as t1
left outer
join `test` as t2
on t1.date
= t2.date
and t2.code = '002'
where t1.code = '001'

union all

select t2.date
, null
, t2.value
from `test` as t1
right outer
join `test` as t2
on t1.date
= t2.date
and t1.code = '001'
where t2.code = '002'
and t1.date is null

you'll need mysql 4.0 to support union, and if you don't have any code 002 rows without a code 001 row, then you can get away with just the first of the two queries|||ah, the good old denormalization questionI wondered what on earth you were babbling about as I read your response, but when I went back and re-read the question carefully I discovered that you were correct! I've always preferred the structure:SELECT date
, Sum(CASE WHEN '001' = code THEN value END) AS '001'
, Sum(CASE WHEN '002' = code THEN value END) AS '002'
, Sum(CASE WHEN '003' = code THEN value END) AS '003'
FROM test
GROUP BY dateThis only has to make one pass through the table, so it can be more efficeient if the optimiser does its job.

-PatP|||i babble often

i'm wrong occasionally

:cool:|||Thank you all very much !! :)
Paolo Saudin