Dear Yuandan,
The iterations are behaving as if 'animal' is equivalent to 'error'.
There appears to be no information to estimate both error and animal.
ASREML says the AI matrix is singular so that 'animal' and 'error'
variance components are confounded.
If you take the function S2*(1 + Gamma) you get
9.1212 9.2489 9.7919 10.811 11.22 11.264 11.268
from the iterations shown.
Suggesting an simple partitioninging of the error variance.
Hence an increase in the 'animal' variance ratio is reflected
in a reduction in the 'error' variance.
There appears to be 398 animals in the pedigree, 207 records of
data giving 199 residual df. So the pedigree is not obviously
deficient but maybe the fixed effects break the few pedigree links
present!
I would expect this behaviou if the model included 'units'
instead of 'animals'
If you can't see something inadequat with the pedigree, you had better
send me the example.
Arthur
> X-Sender: yzhang@metz.une.edu.au
> Date: Mon, 08 Jun 1998 23:06:15 +1000
> To: asreml@chiswick.anprod.CSIRO.AU
> From: Yuandan Zhang <yzhang@metz.une.edu.au>
> Subject: some output from asr file
> Mime-Version: 1.0
>
> Hi,
> While I analysed sheep weaning weight data using asreml, I've getten some
> output that confused me.
> First, the description of the data is showed as
>
> Model term Size Type COL Minimum Mean Maximum #
> zero #miss
> 1 animal 398 Direct 1 3.0000 204.47 398.00
> 0. 0
> 2 sire 1 Covariate 2 1.0000 31311. 94087.
> 0. 2
> 3 dam 1 Covariate 3 90002. 92419. 94288.
> 0. 0
> 4 sex 2 Factor 4 1.0000 1.5314 2.0000
> 0. 0
> 5 brrank 3 Factor 5 1.0000 1.5990 3.0000
> 0. 0
> ......
> 9 genotype 5 Factor 9 1.0000 3.4734 5.0000
> 0. 0
> ......
> 14 wwt 1 Variate 14 10.900 22.447 34.200
> 0. 0
> ......
> 21 mu 1 Constant Term
>
> Following is the as file
>
> Lamb fec and weight data from mating97 FEC -6/3, wt birth-6/3
> animal !P
> sire
> dam
> sex 2
> brrank 3
> sire_gd
> ewegd
> sirebd
> genotype 5
> lambplot
> bage
> wage
> bwt
> wwt
> wt4298
> wt6398
> FEC181297
> Nem181297
> FEC6398
> nem6398
> #97002 94087 93193 2 2 0 0 1 4 11 94 2.6 19.5 21.7 20.8 4 3 3 0
> wt_fec.txt
> wt_fec.txt
> wwt ~mu sex genotype !r animal !f brrank
>
> The output in asr file is :
>
> Forming 410 equations: 9 dense
> NOTICE: 3 (more) singularities,
> LogL=-331.254 S2= 8.292 199 df 0.10000 0.00000 1.00000
> LogL=-331.012 S2= 7.681 199 df 0.20413 0.00000 1.00000
> LogL=-329.910 S2= 5.400 199 df 0.81331 0.00000 1.00000
> LogL=-327.708 S2= 1.568 199 df 5.89486 0.00000 1.00000
> LogL=-326.870 S2= 0.1704 199 df 64.84345 0.00000 1.00000
> LogL=-326.777 S2= 0.1577E-01 199 df 713.27793 0.00000 1.00000
> LogL=-326.768 S2= 0.1436E-02 199 df7846.05719 0.00000 1.00000
>
> Final parameter values 7846.12527 0.00000 1.00000
>
> WARNING: 1 singularities in AI matrix.
> Source Model terms Gamma Component Compnt/StndErr
> animal 398 398 7846.13 11.2667 9.97
> Variance 207 199 1.00000 0.143596E-02 0.00
> Solution Standard Error T-value T-prev
> 9 genotype 4 5.16 5.16 [ DF F_inc F_all ]
> 2 -1.96544 1.87758 -1.05
> 3 -3.62408 2.28853 -1.58 -0.72
> 4 4.42244 1.71869 2.57 3.65
> 5 4.42008 1.74527 2.53 0.00
> 4 sex 1 3.83 3.99 [ DF F_inc F_all ]
> 7 -0.834937 0.417911 -2.00
> 1 animal 398 effects fitted
> 5 brrank 3 effects fitted
> Finished: 22:37:03.03 LogL Converged
>
> And mu in sln file is 0.
>
> Can some one explain the results for me?
> Thanks
> Yuandan
>
>
>
> ************************************************************************
> Yuandan Zhang
>
> Department of Animal Science
> University of New England
> Armidale, NSW 2351, Australia
>
> Phone +61 + 2 + 6773 2756(Lab)
>
>
> Fax +61 + 2 + 6773 3275
> E-mail yzhang@metz.une.edu.au
> homepage http://metz.une.edu.au/~yzhang/
> ************************************************************************
>
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD email: Arthur.Gilmour@agric.nsw.gov.au
Senior 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
ASREML is currently free by anonymous ftp from pub/aar on ftp.res.bbsrc.ac.uk
Point your web browser at ftp://ftp.res.bbsrc.ac.uk/pub/aar/
in the IACR-Rothamsted information system http://www.res.bbsrc.ac.uk/
To join the asreml discussion list, send the message
subscribe
to asreml-request@chiswick.anprod.CSIRO.au
The address for messages to the list is asreml@chiswick.anprod.CSIRO.au
<> <> <> <> <> <> <>
"Seek first the kingdom of God and His righteousness" Jesus; Matthew 6: 33
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>