Polynomial random regression

From: Ericsson Tore <Tore.Ericsson_at_SKOGFORSK.SE>
Date: Tue, 28 Dec 2004 23:24:17 +0100

Dear friends,

I cannot figure out how to specify a polynomial model with time series data arranged different from the cookbook example http://www.uncronopio.org/luis/asreml_cookbook_lon.html#longit: We have increment core ring density data, one record per tree (genetic analysis, with pedigree). There is no time variable, like t1 t2 t3 ... = 1 2 3 ... since there is one ring measured for each year in sequence (some missing data, though). We intend to use the subset 2--11 (spaced 1 yr). Attempt so far, first some background, the problem in !PART 07:
 TID !P # tree id
 BLK 7 # fix effect
# rings 1--15 from bark
# selection: 0 = bad ring, 1 = OK
 s1 s2 s3 s4 s5 s6 s7 s8 s9 s10 s11 s12 s13 s14 s15
# density: convert bad rings to missing
 d1 !*s1 !M0 d2 !*s2 !M0 d3 !*s3 !M0 d4 !*s4 !M0 d5 !*s5 !M0
 d6 !*s6 !M0 d7 !*s7 !M0 d8 !*s8 !M0 d9 !*s9 !M0 d10 !*s10 !M0
 d11 !*s11 !M0 d12 !*s12 !M0 d13 !*s13 !M0 d14 !*s14 !M0 d15 !*s15 !M0
!PART 07 # Random regression
d2 d3 d4 d5 d6 d7 d8 d9 d10 d11 ~ Trait Tr.BLK !r Tr.TID
??? !PVAL 1:10 ??? pol(???,4) ???

Do I have to 'unflatten' the data file and add a year variable?

Thanks for any help and a Happy New Year!


Tore Ericsson, PhD
Skogforsk, Box 3
SE-918 21 Sävar, Sweden
+46(90)2033355 (office)
+46(90)2033360 (fax)
+46(70)3198352 (mobile)
Received on Tue Dec 28 2004 - 23:24:17 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.)