Re: [R] How to extract partial predictions, package mgcv

> So almost every one of the last set of plotting needed one or more > additions > of "f$" or "data=f" in order to run without error. Perhaps you > attached "f" > earlier and didn't include that code? I personally have given up using > attac

Re: [R] How to extract partial predictions, package mgcv

t >> runs without error on my machine: >> >> On Sep 16, 2009, at 12:39 PM, Staffan Roos wrote: >> >>> >>> Thanks Simon, >>> >>> I wasn't aware that "predict.gam" could be used in this situa

Re: [R] How to extract partial predictions, package mgcv

ot;)] names(f) ### This data, i.e. "f", can now be exported or used for further calculations within R #plot the data par(mfrow=c(1,1)) plot(pp_s.x0.~x0,type="p",pch=2,cex=0.50) sequence=order(x0) lines(x0[sequence],pp_s.x0.[sequence]) lines rug(jitter(x0)) ##

[R] How to extract partial predictions, package mgcv

erel egg densities. Using those graphs as an example, I would like to extract the partial predictions for e.g. "s(b.depth)", given the inclusion of the other variables. -- Staffan Roos, PhD Research Ecologist BTO Scotland __ R-help@r-project.