Fwd: Re: overall mean and predictions
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Fwd: Re: overall mean and predictions




>Date: Tue, 29 May 2001 08:35:52 -0500 (CDT)
>From: <southey@ux1.cso.uiuc.edu>
>X-X-Sender:  <southey@staff3.cso.uiuc.edu>
>To: "Clarke, Peter" <PClarke@agric.wa.gov.au>
>cc: "'asreml@chiswick.anprod.csiro.au'" <asreml@chiswick.anprod.csiro.au>
>Subject: Re: overall mean and predictions
>
>Hi,
>It is hard to say what the problem is with the little amount of
>information that you provide. Since you don't describe the GENSTAT model,
>do you get the same result when you use a fixed effects only model or a
>different constraint?
>
>In anycase, I would suggest that you create a .pin file where you specify
>the contrasts necessary to obtain what you want.
>
>Regards
>Bruce
>
>
>On Tue, 29 May 2001, Clarke, Peter wrote:
>
> > Dear ASREMLers,
> >       Can someone please throw some light on the following ASREML
> > prediction:
> >
> > I have a model predicting variety means using code:
> >
> > yield ~ mu c(Varname) c(tos) c(Year) c(Newzone),
> > !r TrialID
> > 1 1 0 !WT=nrep
> > 0 0 0
> > predict Varname
> >
> >  The predicted variety means are all exactly 62 kgs lower than those from
> > GENSTAT.
> >  However, the raw mean is 2347 (which I presume GENSTAT adjusts to,
> >  BUT ASREML estimates mu=2421
> >  So, strangely, the ASREML overall mean is HIGHER???
> >  than GENSTAT.
> >
> >           Peter Clarke
> > --
> > Asreml mailinglist archive: 
> http://www.chiswick.anprod.csiro.au/lists/asreml
> >


--
Andrew Swan
CSIRO Livestock Industries
Locked Bag 1
Armidale NSW 2350
phone: 02 6776 1377
fax: 02 6776 1333

--
Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml