Re: extra variance component

I have been having similar problems.

What would the specifications be for Asreml2, using a model with
y ~ mu fixed1 fixed 2 fixed1:fixed2 !r !{ ran1 fixed1.ran1 !},
                                       !{ide(ran1) fixed1.ide(ran1) !}?

I have had the variance structures set up as:

1 1 4

newid 1
newid 0 ID 0.3 #!GP

sCONES.newid 2
sCONES 0 US !GP
1.5
newid

ide(newid) 1
ide(newid) 0 ID 1.5 !GP

sCONES.ide(newid) 2
sCONES 0 US !GP
1.5
ide(newid)

But this doesn't seem to work.

Thanks,
Melissa

arthur.gilmour_at_DPI.NSW.GOV.AU wrote:
> Dear Alistair,
>
> You wrote
> I have just changed to ASReml 2 and am having a small problem with my
> animal models - specifically I keep getting an extra "variance" term
> reported. I've just run a few simple univariate models and this seems to
> only occur if I actually include the variance model. So for example;
>
> y ~ mu !r animal
>
> Gives me Va and a residual variance as expected but,
>
> y ~ mu !r animal
>
> 1 1 1
>
> 0 0 ID 0.35
>
> ANIMAL 1
> ANIMAL 0 ID 0.02
>
>
> Gives me e.g.
>
> Source Model terms Gamma Component Comp/SE %
> C
> Variance 5571 5562 1.00000 3.37476 0.00
> 0 S
> Residual identity 5571 0.111444 0.376097 39.74
> 0 U
> ANIMAL Ainverse 10044 0.544778E-02 0.183850E-01 2.79
> 0 U
>
> The problem being the extra, unexpected "variance" component.I checked
> the updated part of the manual but couldn't work out what I'm doing
> wrong. Has the coding of the variance model changed or am I just
> missing something obvious? Any thoughts would be very welcome.
>
> Alastair
>
> The model is the same as before, or should be, but is not specified
> correctly
> (the residual variance is implicitly present twice).
> What has changed is that ASReml now stops when it detects a singularity in
> the Average Information matrix so that the problem can be fixed rather
> than
> possibly not being noticed as in ASReml 1.
>
> So the correct ways this model can be written are:
>
> y ~ mu !r animal
> 0 0
>
> or
>
> y ~ mu !r animal
>
> 0 0 1
>
> ANIMAL 1
> ANIMAL 0 ID 0.02
>
> or
>
> y ~ mu !r animal
>
> 1 1 1
> 0 0 ID 0.35 !S2==1
>
> ANIMAL 1
> ANIMAL 0 ID 0.02
>
> or
>
> y ~ mu !r animal
>
> 1 1 1
> 0 0 ID
>
> ANIMAL 1
> ANIMAL 0 ID 0.02
>
> Note that the third of these may fit the variance model on the variance
> scale rather than on the GAMMA scale which is the default for the other
> models.
>
> I trust this clarifies the situation.
>
> ASREML 2 is now available from http://www.VSNi.co.uk
>
> Check out my personal site http://www.cargovale.com.au/
>
> Grace, mercy and peace to you from God our Saviour Jesus Christ
>
> Arthur Gilmour PhD
> mailto:Arthur.Gilmour_at_dpi.nsw.gov.au
> 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
>
> http://www.dpi.nsw.gov.au/reader/17263
> http://www.VSNi.co.uk/products/asreml
> The ASReml discussion group is at ASREML-L_at_dpi.nsw.gov.au
> To join it, mailto:arthur.gilmour_at_dpi.nsw.gov.au
> Cookbook: http://uncronopio.org/ASReml
>
> Proposed travel:
> Aust Genstat Conf 5-8 Dec Victor Harbour
> Leave 18 Dec to 5 Jan
> ASReml workshop (Forestry) 16-18 April Tasmania
> ASReml workshop (Animal Breeding) May South Africa
> AAABG 24-26 September Armidale
> <><><><><><><><><><><><><><><><><><><><><><><><><>
>
> 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 Sun Sep 19 2006 - 14:18:53 EST