R-structure
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

R-structure



Hi ASREMLers!
Any suggestions about these two problems? 

1. A question about the structure of the residual (R).
I have individual data of feed intake of approx. 3000 pigs at the 25 days
3,4,5,6,7, 17-21, 31-35, 45-49 and 59-63. I would like to try heterogeneous
AR and heterogeneous ARMA for these data, so that R(ARH)=
 
a.a a.b.r a.c.r2 a.d.r3 a.e.r4      (~0)         ...
    b.b   b.c.r  b.d.r2 b.e.r3
          c.c    c.d.r  c.e.r2
     (sim)       d.d    d.e.r
                        e.e   
                               f.f f.g.r f.h.r2  ...
                                   g.g   g.h.r   ...
                                         ...     ...

How could I do that? Perhaps using 'POW'? Gamma results using DIAG were:
1.18 0.91 1.00 0.98 1.09
1.21 1.22 1.20 1.22 1.35
1.44 1.45 1.46 1.31 1.35
1.66 1.72 1.67 1.52 1.63
1.96 1.95 1.80 1.64 1.76

2. Is the G matrix correct as follows or should I modify it according to the
data structure?

fi~ mu !r ident ident.x ident.x2 !f mv

0 0 1

ident 2
3 0 US !+6
0.03681
0.001 0.08332
0.001 0.001 0.08332
ident 0 AINV

 
§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§
§ 			                                    §                                 
§  Justo Lorenzo Bermejo                                    §
§  Institut für Tierzucht und Tierhaltung                   §
§  Olshausenstr. 40                                         §  
§  24118 Kiel  	                                            § 
§  Germany                                    &_____        § 
§                                             /     ^..^    § 
§  T.n. 0431  880 7315                        \_____(oo)    §
§  Fax  04340 402 761                          WW WW        §
§  jlorenzo@tierzucht.uni-kiel.de                           §
§                                                           §
§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

          La vida está hecha solo de momentos

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