Re:fitting random regressions
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re:fitting random regressions



Subject: Re: fitting random regressions
To: L.A.Apiolaza@massey.ac.nz
Cc: asreml@chiswick.anprod.csiro,au


Dear Luis,
Some comments follow

> Hi,
> 
> I'm interested into analysing repeated measurements of a progeny test
> using a random regression model, with a econd order polynomial. The data
> structure in the data file is:
> 
> tree mother father block age height
> 301    1      0      1      1   0.6
> 301    1      0      1      2   0.9
> 301    1      0      1      4   1.5
> 301    1      0      1      6   2.6
> 302    2      0      1      1   0.5
> ...
> 
> Reading the data would be:
> 
> Fitting random regression
>  tree    !P
>  mother 198
>  father   1
>  block   10
>  age    height
> rrcam2.dat
> rrcam2.dat
> 
Given the data file is used for the pedigree, this shoul;d give a message
about repeated animals in the pedigree file.  Use the  !REPEAT option on the
pedigree file name line.

You should then be able to add a model
  height ~ mu age !r tree block
  
  and get an anlysis to start with.

> I haven't manage how to specify the model for including the tree and
> block effect as random using a random regression with a second order
> polynomial.  From reading the manual I've got the impression I should
> use !ASMV for specifying a multivariate analysis with univariate data
> presentation and unit for fitting the error. I've tried several
> approaches for the analysis but with no success.  I would appreciate
> very much any suggestions or comments about the coding for the analysis.
> 
The !ASMV qualifier controls the error structure.  
If the data has no missing values, it is not required.
Or, if the  mv  model term is included, it is not required.

Rather than a single error, you want a multivariate error

So now try the model

 height ~ mu age !r tree block !f mv
 1 2
 NNN     # Replace with the number of data records/4
 4 0 US a 0 a 0 0 a 0 0 0 a    # replace A with the error variance obtained
                                 in the previous run.
                                 
 Once you have this running and have updated the starting values for
 for the Error variance,   
 
 Expand the model to the random regressions
 e.g. something like
 
 height ~ pol(age,2) !r tree.pol(age,2) block.ide(age) !f mv
 1 2  2
 NNN     # Replace with the number of data records/4
 4 0 US a 0 a 0 0 a 0 0 0 a    # replace A with the error variance obtained
                               # in the previous run.
                                 
 tree.pol(age,2) 2
 tree 0
 pol(age,2) 0 DIAG .1 .1 .1
 #pol(age,2) 0 US .1 .01 .1 .01 .01 .1   # Use this form when you have SVs
 
 block.ide(age)  2
 block 0
 ide(age) 0 CORR .1 .1 .1 .1 .1
 
 
So, you should perform the analysis in a sequence
    a) a simple model to get the data read and factors checked,
    b) then get the error structure right
    c) then build up to the model you actually want
    
I have not used  ide(age) before to pick up levels generated my the pol
command so this may not work.  If not, insert a new factor which codes
the repeated measures 1 2 3 4  and use it instead.

Let us know how it goes.
Arthur

> Thanks in advance,
> 
> Luis Apiolaza
> Institute of Veterinary, Animal and Biomedical Sciences
> Massey University
> Palmerston North
> New Zealand
> 
> 
> 

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
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

>>>>>>>>>>>>  ASREML WORKSHOP sheduled 10-12 March: $300  <<<<<<<<<<<<<

 But He answered and said, "It is written, `Man shall not live by bread alone,
   but by every word that proceeds from the mouth of God' "  Matthew 4:4
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
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

>>>>>>>>>>>>  ASREML WORKSHOP sheduled 10-12 March: $300  <<<<<<<<<<<<<

 But He answered and said, "It is written, `Man shall not live by bread alone,
   but by every word that proceeds from the mouth of God' "  Matthew 4:4
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>