Calculating and using uniterrors in ASREML-R

From: <Scott.Chapman_at_CSIRO.AU>
Date: Fri, 31 Jul 2009 09:03:58 +1000

ASREML in GenStat allows use of 'uniterror' as part of 2nd stage type analyses. From what I understand these uniterror ('weights') equal:



v = the diagonal from the inverse of the Sigma (G) matrix.

My shortcut to do this has usually been to take the standard errors from the Genotype predictions (and could therefore use BLUEs or BLUPs to do it).
But should be able to do from the asreml object.

1. I did have some code in the 'samm' days that could calculate the ginverse.

    How can I get the uniterrors (for Genotypes) from some of the following types of models?

    b.asreml <- asreml(msg ~ 1, random=~G + sbloc + sbloc:bloc, data=b)

    b2.asreml <- asreml(msg ~ G, random=~ sbloc + sbloc:bloc, data=b)
    b3.asreml.corh <- asreml(msg~E, random=~corh(E):G + at(E):sbloc + at(E):sbloc:bloc, data=all, predict=predict.asreml(classify=list("E:G")))
2. How do I use these errors in 2nd stage analyses? (i.e. if data in 3rd model was BLUEs and weights from a series of analyses with the 2nd model....)

scott c.
Received on Sat Jul 31 2009 - 09:03:58 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.)