On Thu, 13 May 1999, Haja Kadarmideen wrote:
> Dear ASREML members,
> A follow-up on discussion on missing values:
> If ASREML gives a value of zero to missing dependant variable,
> denoted as (*), then can anyone tell me how ASREML differentiates
> binary data (dependant variable coded as either 0 or 1) with some
> dependant variables given the value (*) ? Hopefully, it does distinguish
> between two types of zero's.
Haja,
asreml places a zero in the y variate if the keyword mv is included in
the model formula. This includes an extra dummy variate in the fixed
effects design matrix, one column for each mv, the column is zero except
for a 1 where the mv occurs. This is equivalent to dropping the mv. There
are some computational advantages to this approach, especially for
spatial arrays, in which it is useful to preserve structure to utilize
the separability of the error process, ie so that R = Sig_c x Sig_r.
In other situations it is probably ok to discard the mv.
This happens automatically and doesnt affect binary data, ie data with
real zeros
cheers
brian
> > Thanks
> Haja
>
>
>
>
> --------------------------------------------------
> Dr. Haja Kadarmideen
> Animal Breeding and Genetics Department
> Animal Biology Division
> Scottish Agricultural College (Edinburgh)
> Bush Estates
> Penicuik, Midlothian EH26 0PH
> Scotland, UK
>
> Telephone: +44 0131 535 3246
> Fax : +44 0131 535 3121
>
> e-mail : h.kadarmideen@ed.sac.ac.uk
> --------------------------------------------------
> --
> Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml
>
...............................................................................
Brian Cullis Tel: 02 6938 1855
NSW Agriculture Fax: 02 6938 1809
Wagga Agricultural Institute email: brian.cullis@agric.nsw.gov.au
Pine Gully Rd
WAGGA WAGGA NSW 2650
--
Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml