Re: Using the 'onlyuse' parameter in asreml.predict

From: kiwi <asremlforum_at_VSNI.CO.UK>
Date: Tue, 28 Apr 2009 05:58:20 +0100

Arthur wrote:
> ......
> Predicting RowTree with only='ped(RowTree)'
> should just produce the BLUPs for ped(RowTree), the same as a returned
> in the solutions (coeff).
> Without 'only=', they will include average effects of fixed terms
> in the model. I do not understand why you call this 'heritability'
> though.

It's not what I called 'heritability'. I used an equation involving prediction variances and additional calculations from the Amatrix which I got from Brian's code to do with 'ac' data (at the Lincoln workshop). I didn't want to bore readers with that detail which I think I understand well enough. Just what is included in the prediction calculation has a bearing on variances which have a consequence on the heritability calculation. My attempt to keep it brief didn't make it clear what I was trying to do.

Your reply seems to indicate that averaging over the fixed effects (i.e. omitting the 'only' term) is probably the most appropriate in my case. Block was the only fixed effect.

Thanks for your help.

-------------------- m2f --------------------

Sent using Mail2Forum (

Read this topic online here:

-------------------- m2f --------------------
Received on Wed Apr 28 2009 - 05:58:20 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.)