Re: covariances in glm.c

[John Darrington]

I thought I already replied to this, but it doesn't seem to have 
shown up in the archive.  I'll try again:


So far as I'm aware, the current code should generate the correct
covariance matrix with the following caveat:  There are no interactions (yet),
or more acurately, no interactions with more than one variable.

So your command 
     GLM v0 by v1 v2
         /INTERCEPT = include

has an implicit subcommand:

        /DESIGN = v1 v2.

The default SPSS behaviour has the implicit subcommand

        /DESIGN = v1 v2 v1*v2.

Which makes things more complex - but not much.  We can deal 
with the interactions later, once we have the sums of squares correct.

J'

On Mon, Feb 07, 2011 at 09:08:34PM -0500, Jason Stover wrote:
     I'm testing the patch for type 3 sums of squares in glm.c
     with this syntax:
     
     data list list / v0 to v2.
     begin data
     3.2 1 1
     3.1 1 1
     3.3 1 2
     3.4 1 2
     3.2 1 3
     3.3 1 3
     3.3 1 4
     3.2 1 4
     2.8 2 1
     2.9 2 1
     3.3 2 2
     3.0 2 2
     3.1 2 3
     3.2 2 3
     3.2 2 4
     3.1 2 4
     end data
     GLM v0 by v1 v2
         /INTERCEPT = include.
     
     I'm getting some covariances that don't look right. 
     Is glm.c currently supposed to get the covariances correct
     from such syntax? (Or is there something wrong with the syntax?)
     
     -Jason
     
     _______________________________________________
     pspp-dev mailing list
     address@hidden
     http://lists.gnu.org/mailman/listinfo/pspp-dev

-- 
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://pgp.mit.edu or any PGP keyserver for public key.

signature.asc
Description: Digital signature