RE: ASREML problem
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: ASREML problem



Dear ASREML friends,

Those who have picked up the latest ASREML (27 Aug 99) will notice that
the 'LogL Converged' message now has an alternative form
'WARNING: LogL Converged; Parameters Not Converged'

Both messages mean that the change in LogL was very small
at the last interation and is therefore presumed to have converged.

The second form points to the fact that a change greater than 1%
occurred to at least one parameter on the final update.
To identify which parameter is still changing and how big the change is,
look in the % column (the second last column) of the variance component
table.  At convergence, you expect this column to be all zero.
'Parameters Not Converged' means some are not zero.

You should look at the table and decide whether you want to force ASREML to 
do a few more iterations.  You can do that by rerunning the
job with  -c (or !CONTINUE) to do 2 more iterations.

Or use the new  !EXTRA i   qualifier (on the data file name line)
to request ASREML to do i more iterations after convergence on the LogL value.



 Source                Model  terms     Gamma     Component    Comp/SE   % C
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4  0.100000E-02  0.100000E-02   0.00   0 F
 Residual            CORRelat     4   15.8457       15.8457      71.49   0 P
 Residual            CORRelat     4   17.9434       17.9434      62.43   0 P
 Residual            CORRelat     4   18.4470       18.4470      51.89   0 P
 Residual            CORRelat     4   19.1084       19.1084      41.95   0 P
 pol(ln,0).an        identity     1  0.873328      0.873328       2.71   0 U
                                                                         ^
                                                                         |
                                                                                                        

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD                  mailto:Arthur.Gilmour@agric.nsw.gov.au
Senior Research Scientist (Biometrics)               fax: <61> 2 6391 3899
NSW Agriculture                                           <61> 2 6391 3922
Orange Agricultural Institute             telephone work: <61> 2 6391 3815
Forest Rd, ORANGE, 2800, AUSTRALIA                  home: <61> 2 6362 0046

ASREML is still free by anonymous ftp from pub/aar on ftp.res.bbsrc.ac.uk
    or point your web browser at ftp://ftp.res.bbsrc.ac.uk/pub/aar/ 

To join the asreml discussion list, send the message  
     subscribe
mailto:asreml-request@chiswick.anprod.CSIRO.au

To send messages to the list, mailto:asreml@chiswick.anprod.CSIRO.au

Asreml list archive: http://www.chiswick.anprod.csiro.au/lists/asreml

                        <> <> <> <> <> <> <>
"Christ Jesus came into the world to save sinners"   I Timothy 1:15.
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>


--
Asreml mailinglist archive: http://www.chiswick.anprod.csiro.au/lists/asreml