Hi all,
I am doing nls regression of the Rogers type III equation. However I can't
find good starting values and keep getting the error messages: "Missing
value or an infinity produced when evaluating the model" OR "singular
gradient matrix at initial parameter estimates"
Is there a way to automatically check different starting values and give a
list of values for each parameter? In my model only Th3 is a parameter
with a good estimate (2.5). for v,w and x I would like to check a broad
range of values. Is it possible to do this in R?
Under my script for plots and nls regression and a copy of the values of
my data file.
plot(FR~N0,main="Rogers III")
RogersIII <-
nls(FR~N0*(1-exp(((v+w*FR)/(1+x*FR))*(Th3*FR-24))),start=list(v=0.002,w=0.00075,x=-0.1,Th3=2.5),control=list(maxiter=1000000,minFactor=0.0000000000000000000000000000000000001))
# 24 is tijd in uren
hatac <- predict(RogersIII)
lines(spline(N0,hatac))
summary(RogersIII)
N0 FR
5 4
3 2
5 3
5 4
6 4
5 4
4 2
5 4
6 5
5 3
10 5
14 7
12 6
17 9
10 4
16 8
15 8
14 5
15 6
15 5
19 10
20 11
22 10
21 12
25 12
26 11
25 10
25 12
25 11
24 9
35 12
33 12
35 13
36 12
31 12
37 12
39 13
36 14
35 12
35 10
46 11
44 10
41 12
43 14
45 12
48 11
45 13
45 9
49 12
45 11
56 12
59 15
52 11
54 10
51 12
55 12
61 13
55 12
54 10
55 12
Kind regards,
Met vriendelijke groeten,
Joachim
Don't waste paper! Think about the environment before printing this e-mail
______________________________________
Joachim Audenaert
Adviesdienst Gewasbescherming
Proefcentrum voor Sierteelt
Schaessestraat 18
B-9070 Destelbergen
Belgium
Tel. +32 9 353 94 71
Fax +32 9 353 94 95
E-mail: joachim.audenaert@pcsierteelt.be
www.pcsierteelt.be
______________________________________
[[alternative HTML version deleted]]
