in nnet(), you should add linout = TRUE. The default setting is logistic output.
hth.
On 1/28/07, Aimin Yan <aiminy at iastate.edu>
wrote:> I use neural network to predict a continuous variable( omega in training).
>
> But I get all "1" instead of real value.
>
> Do you know why?
>
> Aimin
>
> Thanks
>
> The following is code
>
> > m.nn.omega <- nnet(omega~aa_three+bas+bcu+aa_ss,
aata=training,size=2)
> # weights: 57
> initial value 97329662.256069
> final value 96367717.444383
> converged
> > pr.nn.train<-predict(m.nn.omega,training,type="raw")
> > head(pr.nn.train)
> [,1]
> 1 1
> 2 1
> 3 1
> 4 1
> 5 1
> 6 1
> > head(training)
> pr aa_three aa_one aa_ss aa_pos aas bas ams bms acu
> bcu omega y index
> 1 1acx ALA A C 1 127.71 0 69.99 0
> -0.2498560 0 79.91470 outward TRUE
> 2 1acx PRO P C 2 68.55 0 55.44 0
> -0.0949008 0 76.60380 outward TRUE
> 3 1acx ALA A E 3 52.72 0 47.82 0
> -0.0396550 0 52.19970 outward TRUE
> 4 1acx PHE F E 4 22.62 0 31.21 0 0.1270330 0
> 169.52500 inward TRUE
> 5 1acx SER S E 5 71.32 0 52.84 0
> -0.1312380 0 7.47528 outward TRUE
> 6 1acx VAL V E 6 12.92 0 22.40 0 0.1728390 0
> 149.09400 inward TRUE
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
--
WenSui Liu
A lousy statistician who happens to know a little programming
(http://spaces.msn.com/statcompute/blog)