Qfyilyi

So I originally create a plot of my data using different points to show different combinations of factors.

plot(m, y, xlab='Body mass (kg)', ylab='Maximum oxygen uptake (litres/min)', 

ylim=c(0.24,6.6), type='n')

points(m[s==0 & v==0], y[s==0 & v==0], pch=16, cex=0.6)

points(m[s==0 & v>0], y[s==0 & v>0], pch=1, cex=0.6)

points(m[s==1 & v==0], y[s==1 & v==0], pch=17, cex=0.6)

points(m[s==1 & v>0], y[s==1 & v>0], pch=2, cex=0.6)

I have then fitted a non-linear model to my data using nls.

fitness0.st <- list(a0=2/3, a1=0, a2=0, a3=0, b0=coefs[1], b1=coefs[2], 

b2=coefs[3], b3=coefs[4], g0=coefs[5], g1=coefs[6], g2=coefs[7], 

g3=coefs[8])

fitness0.nls <- nls(y ~ m^(a0+a1*s+a2*v+a3*sv)*(b0+b1*s+b2*v+b3*sv+ 

(g0+g1*s+g2*v+g3*sv)*a), start = fitness0.st) 

And now I want to consider how well my model fits the data. I have produced the residual plots for my model using

plot(m,residuals(fitness0.nls), xlab='Body mass (kg)', ylab='Residuals', 

xlim=c(30,130))

plot(fitted(fitness0.nls), residuals(fitness0.nls), xlab='Fitted values', 

ylab='Residuals', pch=20)

qqnorm(residuals(fitness0.nls), pch=20, main='Normal QQ-plot')

qqline(residuals(fitness0.nls))

But I would also like to create a pot to show the fit of my model using predict but I am unsure how to do this. Can anyone help?