Re: error: failed while ordering equations

From: Bruce Southey <southey_at_UIUC.EDU>
Date: Thu, 1 May 2008 09:14:20 -0500

ian white wrote:
> Dear Listers
> Not come across this error message before, not in manual. The analysis
> is of a single trait, approx 700 records, a few fixed effects, one
> random effect for a largish pedigree (maybe 10,000 individuals).
> Any ideas what might be cause?
> --
> I White
> University of Edinburgh
> Ashworth Laboratories, West Mains Road
> Edinburgh EH9 3JT (Tel 0131 650 5490)
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
You probably have tried many options but I would test that:
1) The model runs correctly without the random effect and pedigree
declaration as that would imply an error with your pedigree.
2) The basic model with no fixed terms runs correctly as that implies
modeling error.

If it is a pedigree related error, I would suggest:
1) Force it to recompute the A matrix as it could be trying to use an
old version of the A matrix.
2) A real error in the pedigree file that is not caught and may be
something as simple pedigree labeling. Try a smaller pedigree or
recoding the pedigree.
3) A memory related error that may be computer or operating system
related - you could have memory exhaustion so you need another system.

4) A larger problem lurks if there is some type of confounding between
fixed effects as well as between fixed and random effects. That would
mean fitting models of increasing complexity until it breaks.

Hopefully it is a simpler error.

Received on Sat May 01 2008 - 09:14:20 EST

This webpage is part of the ASReml-l discussion list archives 2004-2010. More information on ASReml can be found at the VSN website. This discussion list is now deprecated - please use the VSN forum for discussion on ASReml. (These online archives were generated using the hypermail package.)