[R] Prediction plots

Boris Steipe boris.steipe at utoronto.ca
Wed Apr 19 18:20:34 CEST 2017


Can you provide a small reproducible example and explain what exactly is going wrong?

Just a handful of data points will do.


B.



> On Apr 18, 2017, at 2:16 PM, Santiago Bueno <swbueno at gmail.com> wrote:
> 
> Thanks Boris, the following is an extract of my data. I have developed
> biomass models using codes like:
> 
> start <- coef (lm(log(Btot)~I(log(dbh**2*haut)),data=dat[dat$Btot>0,]))
> 
> start[1] <- exp(start[1])
> 
> names(start) <- c("a","b")
> 
> M1 <- nls(Btot~a*(dbh**2*haut)**b,data=dat,start=start,weights=1/dat$dbh**4)
> 
> 
> start <- coef(lm(log(Btot)~I(log(dbh))+I(log(haut)),data=dat[dat$Btot>0,]))
> 
> start[1] <- exp(start[1])
> 
> names(start) <- c("a","b1","b2")
> 
> M2 <- nls(Btot~a*dbh**b1*haut**b2,data=dat,start=start,weights=1/dat$dbh**4)
> 
> 
> Tree No dbh haut Btot
> 1 35.00 18.90 0.535
> 2 25.00 16.60 0.248
> 3 23.00 19.50 0.228
> 4 13.50 15.60 0.080
> 5 20.00 18.80 0.172
> 6 23.00 17.40 0.190
> 7 29.00 19.90 0.559
> 8 17.60 18.20 0.117
> 9 31.00 25.30 0.645
> 10 26.00 23.50 0.394
> 11 13.00 13.00 0.038
> 12 32.00 20.70 0.443
> It is my interest to get prediction plots for the models. I have tried to
> use the following codes with no success: Let m be one of the fitted models
> with dbh as the only entry. To construct a plot of the predictions made by
> this model I have tried:
> with(dat,plot(dbh,Btot,xlab="Dbh(cm)",ylab="Biomass (t)"))
> D <- seq(par("usr")[1],par("usr")[2],length=200)
> lines(D,predict(m,newdata=data.frame(dbh=D)),col="red")
> For a model m that has dbh and height as entries, I have tried to get its
> predictions as follows:
> D <- seq(0,180,length=20)
> H <- seq(0,61,length=20)
> B <- matrix(predict(m,newdata=expand.grid(dbh=D,height=H)),length(D))
> 
> Can someone provide help please!!!
> 
> 
> Best regards,
> 
> Santiago Bueno
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.



More information about the R-help mailing list