Re: ASREML Augmented Designs problem
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: ASREML Augmented Designs problem



> > From: Vincenzo Matassa <s185152@student.uq.edu.au>
> To: Arthur Gilmour <gilmoua@ornsun.agric.nsw.gov.au>
> Subject: ASREML Augmented Designs problem
> 
> Dear Asremlers,
> 	I am baffled in another problem of mine. Pretend we have an
> Augmented randomised complete block experiment. That is a randomised
> complete block design with unreplicated genotypes augmenetd in the
> design. For instance:
> 
> 		Rep1	Rep2
> 
>                 check   5 
> 		1       check
> 		2       6
> 		3       7   
> 		4       check        
> 		check   8 
>  
> Now if we think of the linear model;
> 
> 	Yield= grand_mean + Block effects + Genotype +Check + Error    (1)
> 
> were all main effects are considered fixed.
> 
> The Fitted value for say Genotype (1) should be;
> 
>         Yield(1)=rep1 effect + raw genotype(1) yield value        (2)
> 
> If I fit 
> 
> 	Yield ~ mu rep check  genotype                   (3)
> 
> in ASREML.The fitted value for Genotypoe (1) say is just the raw yield
> value. I believe it should be the form given in eq.(2).
> 

QUESTION 1: how is 'genotype' coded in the CHECK plots?
If we assume genotype is 0 for check plots
and CHECK is coded a particular value for all TEST plots.


Without constraints or random effects
   rep_1 effect is zero;
   check_1 effect is zero and 
   genotype_1 effect is zero


However; if CHECK is coded 0 for all TEST plots,
then NO singularity occures in the GENOTYPE factor
so  Genotype_1 will be the deviation from the mean.



X matrix becomes
   mu r1 r2  C g1 g2 g3 g4 g5 g6 g7 g8
    1  1  0  1  0  0  0  0  0  0  0  0
    1  1  0  0  1  0  0  0  0  0  0  0
    1  1  0  0  0  1  0  0  0  0  0  0
    1  1  0  0  0  0  1  0  0  0  0  0
    1  1  0  0  0  0  0  1  0  0  0  0
    1  1  0  1  0  0  0  0  0  0  0  0
  
    1  0  1  0  0  0  0  0  1  0  0  0
    1  0  1  1  0  0  0  0  0  0  0  0
    1  0  1  0  0  0  0  0  0  1  0  0
    1  0  1  0  0  0  0  0  0  0  1  0
    1  0  1  1  0  0  0  0  0  0  0  0
    1  0  1  0  0  0  0  0  0  0  0  1

       S        S
Here, the first REP effect is singular and the first genotype will
be singular becasue the Check and Genotype combine to make MU

If terms were fitted in a different order in ASREML; so that MU and
REP_1 were singular, then  you would get the effect you describe.

CHeck the .sln file to see what is singular.

Arthur   
> I am most grateful for this discussion group. I thank you, once again for
> your advice.
> 
> Regards Vince
> 
>