Thanks for your suggestions, however, the .ASR file and .VVP file don't
indicate the problems you have suggested (please find attached plus the .VRB
file).
I was suspicious that this was a problem with data structure being inadeqate
for the model, but found no evidence of the usual problems that would
suggest this (e.g. failure to converge, obvious problems with estimates of
variance components etc). I have only just started using ASREML and was
therefore unsure whether it was me, the data or the program that was causing
the problem! Unfortunately I am still not sure.
I have started with a simple animal model and built up from there. I have
also used the same models for analysis of other data sets that are very
similar (but for a different breed). Some data sets which I believed to be
marginal in terms of estimating maternal effects have been analysed without
giving this error, in contrast I expected this one to be better for the
estimation of maternal effects as it has reasonably good depth of pedigree
and a relatively high proportion of dams have their own performance records.
Once again many thanks for your help
-----Original Message-----
From: bsouthey@iastate.edu <bsouthey@iastate.edu>
To: Janet Roden <jjr997@aber.ac.uk>
Cc: asreml <asreml@chiswick.anprod.csiro.au>; bsouthey@iastate.edu
<bsouthey@iastate.edu>
Date: 05 January 2000 22:48
Subject: Re: Fortran error
>Ok,
>Now I know what you trying to do.
>
>Check the .asr file for messages like:
>Fault 0 Convergence Failed
>Warning: Code B - fixed at boundary
>
>Also is the Compnt/StndErr zero for your variance components?
>
>Look at the .vvp file, it is probably full of zeros.
>
>These indicate problems with the model and/or the data. There is probably
not
>sufficient information in the data to fit the model you want to.
>
>Regards Bruce
>
>