Re: constraint of factor

From: <arthur.gilmour_at_DPI.NSW.GOV.AU>
Date: Thu, 22 Dec 2005 16:45:51 +1100

Dear Sansak,

I could not get asreml1.10 to constrain the fixed effect of contemporary
group (HYS) successfully in my analyses (by fitting c(HYS) in the model).
Asreml still always puts mean of one of the HYS levels to zero.
Would this be because I have too many levels of HYS? if so, up to how many
levels of an effect can we constrain (sum to zero)?
By the way, for effects that do not have many levels, the constraint
command was working fine in my analyses using asreml1.10.

Several things might be happening. c(HYS) will only work as intended if
HYS has the correct number of levels
and there are no missing levels. i.e. If HYS is declared with 100
levels and there are only 99 actually present,
a singularity will be generated.

There is a limit to the number of 'dense' equations of 800, so if HYS has
more than about 800 levels (depending on what else is
in the dense portion), it will be moved into the 'sparse' set. This
should not stop the c() mechanism working though.

Using c() results the the SSP being much more dense and so if HYS is
large, generates much more work for ASReml.
It is generally better to use PREDICT to obtain the means you want rather
than try and get them as effects from the model.


I expect VSN to start distributing Release 2 of ASReml early in the New
Year. (I have previously said November 2005
but that has passed.) If you would like to try it (for the purpose of
giving feed back, and are paying for support), you could ask
VSN for access to the current beta version. It has a different licensing
module which complicates the process slightly.

 Check out
for my personal greetings and update (I don't know what year it is!).

Thanking God for His mercy through 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
The ASReml discussion group is at
To join it,

Proposed travel:
REML Mixed model RR 19Jan
Wool meets meat 22-23 Feb Orange
IBC Quebec 16-21 July 2006
WCGALP8 13-18 Aug Brazil

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 Wed Dec 22 2005 - 16:45:51 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.)