Re: Singularity in Average Information Matrix

From: <arthur.gilmour_at_DPI.NSW.GOV.AU>
Date: Mon, 11 Dec 2006 12:08:30 +1100

Dear Moshen,

You did not include the .giv file so I can't rerun your job.

You reported an iteration sequence

Initial updates will be shrunk by factor 0.224
1 LogL=-43.4186 S2= 613.31 11 df 1.000 0.3000 0.1000
2 LogL=-41.7362 S2= 178.33 11 df 1.000 2.890 0.5409

3 LogL=-40.8129 S2= 19.986 11 df : 2 components constrained
4 LogL=-40.6964 S2= 1.2961 11 df : 2 components constrained
5 LogL=-40.6888 S2= 0.82104E-01 11 df : 2 components constrained
6 LogL=-40.6883 S2= 0.51932E-02 11 df : 1 components constrained

which indicates the residual variance becoming small so this is likely the
The AI singularity error did not occur in the first iteration so the
problem is not
structural but numeric.

You could test this by adding some extra residual variance i.e.

  phenotype !NORMAL 0.5

With a very small data set, it is not unlikely that estimated parameters
want to
move outside the parameter space, particularly when using an animal model.

I trust this helps.

Grace, mercy and peace to you from God our Saviour Jesus Christ

Arthur Gilmour PhD
Principal Research Scientist (Biometrics)
NSW Department of Primary Industries
Orange Agricultural Institute, Forest Rd, ORANGE, 2800, AUSTRALIA

fax: 02 6391 3899; 02 6391 3922 Australia +61
telephone work: 02 6391 3815; home: 02 6364 3288; mobile: 0438 251 426
Personal website:

ASREML 2 is now available from
The ASReml discussion group is at
To join it,

Proposed travel:
Aust Genstat Conf 5-8 Dec Victor Harbour
Leave 18 Dec to 5 Jan
ASReml workshop (Forestry) 16-18 April Tasmania
ASReml workshop (Animal Breeding) May South Africa
AAABG 24-26 September Armidale

This message is intended for the addressee named and may contain
confidential information. If you are not the intended recipient or
received it in error, please delete the message and notify sender. Views
expressed are those of the individual sender and are not necessarily the
views of their organisation.
Received on Sat Dec 11 2006 - 12:08:30 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.)