On Jan 19, 2006, at 21:03 , Sean Davis wrote:
> I have a table that stores data like so:
>
> Create table raw_vals (
> expt_id int,
> clone_idx int,
> val numeric,
> primary key (expt_id,clone_idx)
> );
>
> And I would like to design a query that gives:
>
> Clone_idx expt_id_1 expt_id_2 ....
> 1 0.7834 0.8231 ....
> 2 0.2832 1.2783 ....
>
> There are several tens-of-thousands of clone_idx entries, so doing
> this
> quickly may not even be possible. Any suggestions on how to go
> from this
> normalized structure to the denormalized form easily. Note that
> this isn't
> exactly a crosstab, since there is no counting data, just reshaping.
As far as I know, the crosstab functions in contrib/tablefunc will
still do what you want. Another way is to have as many (self) joins
as you want columns.
Michael Glaesemann
grzm myrealbox com