Re: OWN structures

Hi
I tried some time ago to fit a diagonal error variance which was a smooth
function of time. Even after specifying the OWN structure with a diagonal
sparse inverse I had size problems which I could not understand unless the
matrix is treated as dense.
--------------------------------------------------------
David L. Johnson
Livestock Improvement Corporation
Private Bag 3016
Hamilton, New Zealand
Ph +64-7-8560629
Fax +64-7-8560640
--------------------------------------------------------

             "i.m.s.white"
             <iwhite_at_STAFFMAIL
             .ED.AC.UK> To
             Sent by: ASReml ASREML-L_at_AGRIC.NSW.GOV.AU
             users discussion cc
             group
             <ASREML-L_at_AGRIC.N Subject
             SW.GOV.AU> OWN structures

             21/12/2005 01:18

             Please respond to
               ASReml users
             discussion group
             <ASREML-L_at_AGRIC.N
                SW.GOV.AU>

ASRemlers,

There seems to be a max size to the order of OWN structures. Does
anybody know what it is?

With the structure specified in sparse format, is it feasible to create
a diagonal OWN structure of order approx 10000?

--
************************************************
*    I.White                                   *
*    University of Edinburgh                   *
*    Ashworth Laboratories, West Mains Road    *
*    Edinburgh EH9 3JT                         *
*    Fax: 0131 650 6564   Tel: 0131 650 5490   *
*    E-mail: i.m.s.white_at_ed.ac.uk              *
************************************************