I have had two replies which have helped me sort this out.
Ian White [mailto:imsw@holyrood.ed.ac.uk] suggested that I needed !I on
the animal line.
This is correct, but there is a bit more to it. I also need to specify
the number of levels (74 = 65 progeny as in the data file, + 9 sires),
and I need to modify the data file so that all animals have their own
line (i.e. add the sires), and these need to be in the order implied by
the giv file (in this case each sire must be added just before its first
progeny). The rows & columns of the giv file could be recoded if the
data file was presented with animals in a different order (e.g. all
sires first).
Luis A. Apiolaza [mailto:Luis.Apiolaza@utas.edu.au] gave an easier
solution. This was to leave the animal as a pedigree factor (i.e. keep
the !P), and read in the pedigree as well as the giv file, i.e.
...
ADAILYGAIN
ex11a.dat #to read the pedigree
ex11a.giv
ex11a.dat
AD ~ mu L !r giv(ani,1) 0.25
In this case, the animal factor will be set up correctly. Even though
the ginverse is calculated (or retreived from ainverse.bin), it is not
used, since the giv(ani,1) term forces ASREML to use the ginverse that
is read in.
Thanks Ian and Luis.
Ken Dodds.
-----Original Message-----
From: Dodds, Ken
Sent: Wednesday, November 07, 2001 11:15 AM
To: ASREML message (E-mail)
Subject: Using giv files
I am wanting to use a .giv file with ASREML and decided I would first
try with an example that I could also do normally. To do this I have
used the example file ex11a distributed with ASREML.
So I ran this, but adding the qualifier !GIV to the pedigree file line.
I then renamed asreml.giv to ex11a.giv
Then I created ex11b.as as a copy of ex11a.as, but
removed the !P from the animal data field definition
changed the pedigree file to the giv file
changed ani in the model to giv(ani,1)
----------------- ex11b.as --------------------------
ABSORPTION AND VARIANCE COMPONENTS
animal
sire 9
dam
LINES 2
DAMAGE
ADAILYGAIN
ex11a.giv
ex11a.dat
AD ~ mu L !r giv(ani,1) 0.25
0 0 0
The asr file obtained was:
----------------- ex11b.asr --------------------------
ASReml [20 Sep 2001] ABSORPTION AND VARIANCE COMPONENTS
07 Nov 2001 10:43:52.440 8.00 Mbyte MSWIN ex11b
Your ASReml license expires 31 Mar 2001
* ASREML 2001 *********************************************
* Jan Scale of spline and pol design matrices has changed.*
* This will change the components but not the fit. *
* Feb predict syntax has changed. *
***************************************************** ARG *
Reading ex11a.giv skipping 0 header lines
Inverse G structure of 74 rows having 139 non zero cells read
from ex11a.giv
Reading ex11a.dat FREE FORMAT skipping 0 lines
Univariate analysis of ADAILYGAIN
Using 65 records of 65 read
Model term Size Type COL Minimum Mean Maximum
#zero #miss
1 animal 1 Covariat 1 101.0 133.0 165.0
0 0
2 sire 9 Factor 2 1 5.0154 9
0 0
3 dam 1 Covariat 3 0.000 0.000 0.000
65 0
4 LINES 3 Factor 4 1 2.1231 3
0 0
5 DAMAGE 1 Covariat 5 3.000 4.385 5.000
0 0
6 ADAILYGAIN 1 Variate 6 144.0 176.6 206.0
0 0
7 mu 1 Constant Term
8 giv(ani,1) 1 ex11a.gi 1 101 133.0000 165
0 0
Fault 1 Define structure for giv(ani,1)
Last line read was: 0 0 0 0 0 0 0
ex11b ex11a.dat
Model specification: TERM LEVELS GAMMAS
LINES 3 0.000
mu 1 0.000
giv(ani,1) 1 0.250
SECTIONS 65 4 1
STRUCT 65 0 0 0 0 0 0
8 factors defined [max 500].
4 variance parameters [max 900]. 3 special structures
Final parameter values 0.25000 1.0000
Last line read was: 0 0 0 0 0 0 0
8 1 65 65 1000
Finished: 07 Nov 2001 10:43:52.440 Define structure for giv(ani,1)
Any ideas about what I am doing wrong?
Thanks,
_____________________________________________________________________
Ken Dodds
AgResearch
Invermay Agricultural Centre
Private Bag 50034
Mosgiel
New Zealand
Ph: +64 3489-9083
Fax: +64 3489-9037
Email: ken.dodds@agresearch.co.nz
--
Asreml mailinglist archive:
http://www.chiswick.anprod.csiro.au/lists/asreml
=======================================================================
Attention: The information contained in this message and/or attachments
from AgResearch Limited is intended only for the persons or entities
to which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipients is prohibited by AgResearch
Limited. If you have received this message in error, please notify the
sender immediately.
=======================================================================
--
Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml