Re: Problem using predict
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problem using predict




> 
> Hi,
> 
> I want to get predictions for the subrace effects for four different
> sites (traits), using the following code:
> 
> !PART 21 #Full multivariate Four sites
> !maxit 12
> massflow bawdflow shalflow sso3flow ~ Trait Trait.subrace at(Trait,1).rep,
>                              at(Trait,2).bigrep at(Trait,3).rep,
>                              !r at(Trait,2).plot 38 -at(Trait,3).rep.block,
>                              Trait.csiroID !f mv
> 1 2 1 !STEP .1
> 0 0 0
> Trait 0 US !+10 !GUFUFFUFFFP
>           734
>             0 346
>             0   0 192
>             0   0   0 913
> 
> Trait.csiroID 2
> Trait 0 CORR !+10 !GUUUFFUUUUU
>            1.08
>            0.89 0.68
>            0    0    1.11
>            267  335  174  245
> csiroID
> 
> predict subrace
> 
> The run converges but it does not produce a .pvs file with the
> predictions for subrace, although there are solutions for Trait and
> Trait.subrace in the .sln file.
> 
> The .as file says:
>  Forming  6505  equations:  218  dense
>  Initial updates will be shrunk by factor    0.100
>  Prediction of at(Trait,3)                        1   0   0   -13     0    -1  
   3 not yet enabled
>  Prediction of at(Trait,2)                        1   0   0   -13     0    -1  
   2 not yet enabled
>  Prediction of at(Trait,1)                        1   0   0   -13     0    -1  
   1 not yet enabled
>  Prediction of at(Trait,2)                        1   0   0   -13     0    -1  
   2 not yet enabled
>  Prediction of mv_estimates                    3922   0   0    -4     0     0  
   0 not yet enabled
>  NOTICE:  4540 (more) singularities,
>  LogL=-4165.24     S2=  1.0000       5063 df 
>  LogL=-4163.48     S2=  1.0000       5063 df 
>  LogL=-4159.68     S2=  1.0000       5063 df 
>  LogL=-4157.94     S2=  1.0000       5063 df 
>  LogL=-4157.71     S2=  1.0000       5063 df 
>  LogL=-4157.71     S2=  1.0000       5063 df 
>  Unable to predict at(Trait,3)             4    0    0    1  -13    0   -1    
3
>  LogL=-4157.71     S2=  1.0000       5063 df 
> 
> Not all subraces are in all sites. There are 11, 11, 18 and 8 in each
> site respectively. What is the problem?
> 
> 
> Luis
> 

Dear Luis,

The problem probably is that you will need to be more specific
about the prediction.

The fixed terms in the model include Trait, subrace, rep and bigrep
and mv.

predict subrace 
requests ASREML to average over the other 4 dimensions of a 5way
table but I gather this table has many gaps so the average is undefined.

If you made the rep and bigrep terms random and used the statement

predict Trait subrace

you will have more chance of getting subrace predictions for each Trait
(those where the subrace occurs for that trait).

If you want to keep the current model, try the predict command

predict Trait subrace !present Trait subrace rep bigrep

which should form a Trait x subrace table averaging over rep and bigrep.

This however may not work because rep and bigrep only appear in an interaction
as indeed subrace and ASREML does not yet know how to extract prescence 
information
from an interaction.

Arthur

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD                  mailto:Arthur.Gilmour@agric.nsw.gov.au
Principal Research Scientist (Biometrics)            fax: <61> 2 6391 3899
NSW Agriculture                                           <61> 2 6391 3922
Orange Agricultural Institute             telephone work: <61> 2 6391 3815
Forest Rd, ORANGE, 2800, AUSTRALIA                  home: <61> 2 6362 0046
                                                   Cargo: <61> 2 6364 3288
I havn't finished building my house at Cargo but am back at work.

ASREML is still free by anonymous ftp from pub/aar on ftp.res.bbsrc.ac.uk
    or point your web browser at ftp://ftp.res.bbsrc.ac.uk/pub/aar/ 

To join the asreml discussion list, send the message  
     subscribe
mailto:asreml-request@chiswick.anprod.CSIRO.au

To send messages to the list, mailto:asreml@chiswick.anprod.CSIRO.au

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

                        <> <> <> <> <> <> <>
"Blessed are the peacemakers,
        for they shall be called sons of God"   Jesus; Matthew 5:9

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>





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