If you replace your last plot! statement (before savefig) with:
plot!(pred.x, pred.y, ribbon=(fill(0.3,length(pred.x)),0.5.+0.1*cos.(10pi*pred.x)),fc=:orange,fa=0.3,label="model", linewidth=3)
you get:
OK – my example of
ribbon argument is not realistic, but then I don’t know how to use GLM… If GLM lets you produce prediction confidence intervals, my example should give an idea of how you can do it.