From: | Richard Huxton <dev(at)archonet(dot)com> |
---|---|
To: | "Evandro's mailing lists (Please, don't send personal messages to this address)" <listasjr(at)gmail(dot)com> |
Cc: | pgsql-general <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Beyond the 1600 columns limit on windows |
Date: | 2005-11-08 17:40:14 |
Message-ID: | 4370E2FE.8050708@archonet.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Evandro's mailing lists (Please, don't send personal messages to this
address) wrote:
> Sorry,
> It has nothing to do with normalisation. It is a program for scientific
> applications.
It has everything to do with normalisation. You appear to be pushing
application presentation issues into the structure of your database. If
SQL allowed you, this would break 1NF.
> Data values are broken into column to allow multiple linear regression and
> multivariate regression trees computations.
Sounds like you want an array then (or perhaps several arrays).
> Even SPSS the most well-known statistic sw uses the same approach and data
> structure that my software uses.
Ah - and they've made a good choice?
> Probably I should use another data structure but would not be as eficient
> and practical as the one I use now.
The structure you use inside your application and the data definition
used by the database are two separate things. You presumably are doing
some transformation of data on fetching it anyway - I'd switch
rows-columns over then.
--
Richard Huxton
Archonet Ltd
From | Date | Subject | |
---|---|---|---|
Next Message | Bruno Wolff III | 2005-11-08 17:44:25 | Re: Beyond the 1600 columns limit on windows |
Previous Message | Bruno Wolff III | 2005-11-08 17:35:26 | Re: ident client authentication |