Displaying 1 result from an estimated 1 matches for "modelsimpl".
Did you mean:
modelimpl
2012 Sep 11
1
plotting smoother function on raw data
...tervals.
I <- order(mydata$Ncell)
lines(mydata$Ncell[I], pred.gam$fit[I], lwd = 2)
lines(mydata$Ncell[I], pred.gam$fit[I] + 2 * pred.gam$se.fit[I], lty = 2,
lwd = 2)
lines(mydata$Ncell[I], pred.gam$fit[I] - 2 * pred.gam$se.fit[I], lty = 2,
lwd = 2)
It all works fine if the model instead is:
ModelSimple <- gam(var ~ s(var1, k=4), data = mydata)
However, I get different results (for obvious reasons) if I predict from
Model or ModelSimple.
Please, help me!!!!
Anna
Anna Zakrisson Braeunlich
PhD Student
Department of Systems Ecology
Stockholm University
Svante Arrheniusv. 21A
SE-106 91 Stockh...