>From MAILER-DAEMON@agric.nsw.gov.au Mon Mar 16 15:27:32 1998
Return-Path: <gilmoua@ornsun.agric.nsw.gov.au>
Received: (from bin@localhost)
by spock.agric.nsw.gov.au (8.8.6/8.8.6) id PAA03568;
Mon, 16 Mar 1998 15:26:26 +1100 (EST)
Received: from ornsun(148.145.96.2) by spock via smap (V1.3)
id sma003549; Mon Mar 16 15:25:57 1998
Received: from ornsun by ornsun.agric.nsw.gov.au (SMI-8.6/SMI-SVR4)
id PAA13215; Mon, 16 Mar 1998 15:24:17 +1100
Message-Id: <199803160424.PAA13215@ornsun.agric.nsw.gov.au>
Date: Mon, 16 Mar 1998 15:24:17 +1100 (EST)
From: Arthur Gilmour <gilmoua@ornsun.agric.nsw.gov.au>
Reply-To: Arthur Gilmour <gilmoua@ornsun.agric.nsw.gov.au>
Subject: Re: frequency of singularities
To: richardm@calm.wa.gov.au
Cc: asreml@chiswick.anprod.csiroau
MIME-Version: 1.0
Content-Type: TEXT/plain; charset=us-ascii
Content-MD5: 4wlqLZ549U+l8NCbJGNBsA==
X-Mailer: dtmail 1.2.0 CDE Version 1.2 SunOS 5.6 sun4m sparc
Hi,
> I am interested /concerned about numbers of singularities occuring in
> different analyses. It seems that factors which affect the number of
> singularities include data balance and model specification.
>
> Can any one enlighten me as to how critical it is to eliminate
> singularities, or how many is too many?
> i.e is there a general guideline?
>
> thanks
>
> Richard
>
Dear Richard,
Singularities are not necessarily a problem and it is generally not
necessary to eliminate them if they occur in the first iteration.
If they occur after the first iteration, it is a problem with the
model - maybe some weights have become so small that terms
which were previously included are now considered singular.
You can identify which effects are singular by looking in the .sln file.
Solutions which are zero and have zero for their standard error are
the singular ones. You should work out why each one is singular:
either there is no data for that effect or that there is a linear
dependence. Typically, the first level of a factor will be
singular if the mean has already been fitted. It is generally
not safe to use the 'F_all' F ratio for testing fixed effects if
singularities occur in terms higher in the .asr file.
The effect of a singularirty is to delete the equation in question which
is all you are doing if ytou manually excise singulaitites. Thus,
provided you are happy and can explain each singularity, you may as well
let ASREML handle them.
However, using the con() function will reduce the number of singularities.
Arthur
>
>
> Hi,
> I am interested /concerned about numbers of singularities occuring in
> different analyses. It seems that factors which affect the number of
> singularities include data balance and model specification.
>
> Can any one enlighten me as to how critical it is to eliminate
> singularities, or how many is too many?
> i.e is there a general guideline?
>
> thanks
>
> Richard
>
>
> ======================
> Richard Mazanec
> Dept of Conservation and Land Management,
> Science and Information Division,
> Brain Street,
> Manjimup. 6258
> Phone 08 9771 7985 Int. +61 8 9771 7985
> Fax 08 9777 1183 Int +61 8 9777 1183
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD email: Arthur.Gilmour@agric.nsw.gov.au
Senior Research Scientist (Biometrics) fax: <61> 2 6391 3899
NSW Agriculture telephone work: <61> 2 6391 3815
Orange Agricultural Institute home: <61> 2 6362 0046
Forest Rd, ORANGE, 2800, AUSTRALIA
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
<> <> <> <> <> <> <>
"Blessed are the pure in heart for they shall see God" Jesus; Matthew 5:8
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD email: Arthur.Gilmour@agric.nsw.gov.au
Senior Research Scientist (Biometrics) fax: <61> 2 6391 3899
NSW Agriculture telephone work: <61> 2 6391 3815
Orange Agricultural Institute home: <61> 2 6362 0046
Forest Rd, ORANGE, 2800, AUSTRALIA
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
<> <> <> <> <> <> <>
"Blessed are the pure in heart for they shall see God" Jesus; Matthew 5:8
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>