Re: Problem with ASREML list

# Re: Problem with ASREML list

```Dear All,

!spline and !splpoints are equivalent because ASREML just checks the
first 3 characters of long qualifiers.  Just make sure you no not change CASE.

Concerning how the !splpoints p   argument is used, I offer the following.

ASREML uses p as follows:

It calculates a step size  as range/p

It then sorts the unique values and combines any that are within
the step size.

So  if you have  unique values  1 2 3.5 4 6 7 8 9
p=10  gives step of 0.8   ==>  1 2  3.75 6 7 8 9    i.e. 7 knot points
p=9   gives step of 0.889 ==>  1 2  3.75 6 7 8 9    i.e. 7 knot points
p=8   gives step of 1.0   ==>  1 2  3.75 6 7 8 9    i.e. 7 knot points
p=7   gives step of 1.14  ==>  1.5  3.75 6.5 8.5    i.e. 4 knot points

When points are combined, the reported not value is the simple average
of the unique values (not weighted for their frequency).

On this basis, I expect if your unique values were equally spaced,
(say 1,2,3,..10) and you set p to 9  so that step was 1.0
you would actually get 10 knot points!

So, yes Nick, you just have to accept what ASREML gives because
the distribution of the unique values affects the number of knots as well as
the p value.

Arthur

> Date: Thu, 10 Dec 1998 13:52:42 +0800
> X-Sender: ngalwey@cyllene.uwa.edu.au
> Mime-Version: 1.0
> To: andrew@ram.chiswick.anprod.csiro.au
> From: "N.W. Galwey" <ngalwey@cyllene.uwa.edu.au>
> Subject: Problem with ASREML list
> X-Mozilla-Status2: 00000000
>
> The appended message just got bounced by the ASREML list.   Any chance you
> could help me to send it to a good home?
>
> ___________________________________________
>
>         I am trying to choose the number of knot points for splines, and
> getting some unexpected results.
>
>         I proceed as follows.   I specify a low number of knot points (say
> 4) using the qualifier
>
> !spline 4
>
> in the data file specification in the .as file.   I then write a .pin file
> and generate a .pvs file.   This contains 4 values for each fitted curve, as
> expected.
>
>         However, if I specify a higher number of knot points (say 10), the
> number of fitted values per curve is one or two less than the number
specified.
>
>         The manual cautiously states that the qualifier !SPLPOINTS s affects
> the number of distinct points recotnised by the spl(.) model function - not
> that s is the actual number of knots.   Also, the qualifier given in the
> manual is !SPLPOINTS, whereas I am using !spline - I think I got the latter
> direct from Arthur.   Are they equivalent?   (I can't find anything about
> !spline in the manual.)
>
>         Do I just have to get the precise number of knot points that ASREML
> has chosen from the output?
>
> Nick Galwey
> _____________________________________________________________________
> N.W. Galwey,
> Faculty of Agriculture,
> University of Western Australia,
> Nedlands, WA 6709, Australia.
>
> Tel.: +61 9 380 1959 (direct line)
>       +61 9 380 2554 (switchboard)
> Fax:  +61 9 380 1108

<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>
Arthur Gilmour PhD                    email: 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

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

<> <> <> <> <> <> <>
"For the wages of sin is death,
but the gift of God is eternal life in Christ Jesus our Lord." Romans 6:23
<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>

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