pspp-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: criteria for version 1.0


From: Jason Stover
Subject: Re: criteria for version 1.0
Date: Thu, 16 Oct 2008 16:35:40 -0400
User-agent: Mutt/1.5.18 (2008-05-17)

On Thu, Oct 16, 2008 at 03:20:13PM +0800, John Darrington wrote:
>      Does anyone want to say what they would consider requirements for
>      a 1.0 release?
> 
> 
> 1.  Something to do within subjects analysis of variance and general
>     linear models.   That means, at least the more common parts of the 
>     GLM command, and the ANOVA and MANOVA commands.   I think Jason
>     is doing good work towards this.
> 
> 2.  A way to get good quality tables and charts,  displayed in the
>     GUI, prefereably with some sort of interaction, and to be able to
>     export them to OpenOffice.Org I understand you're working towards
>     that.

I agree that both of these should be the main goals for a 1.0
release. I'm starting to use PSPP in my intro. classes, and it does
the basic analyses acceptably. But the default output is just ugly,
and that's going to turn away a lot of users. Also, GLM, with all its
linear model variants, is the procedure that will do most of the
users' routine analyses, so that should definitely be in a 1.0
release.

About GLM: I think glm.q is going to grow large enough to make glm.q
illegible. It has to be able to do a lot of different tasks, and they
should be split among different files.

So maybe we should all hack on the GLM procedure. The guts of it are
just least-squares, but translating from the user's syntax to the
particular covariance matrix and back will be a lot of work.  I can
handle the least-squares coding and the post-fit computations, but I
think someone else should write the code to read the syntax, dole out
the work to the linear model code, and organize the output. I don't
mind doing it all myself if no one else wants to pitch in, but it will
take a lot longer that way.

I say all this after having spent a lot of time on that one-pass
covariance algorithm, and thinking about all that still needs to be
done in src/math/ to support GLM.

What do you think?

-Jason




reply via email to

[Prev in Thread] Current Thread [Next in Thread]