Alright, I apologize for this basic question - I am both an R and loess
noob.
I am trying to predict the values of column Y in data1 (100000x18 entries)
using a loess fit on training (500x18 entries) and columns A B and C. (training
are not members of data1)
fit <- loess(Y ~ A + B + C, training)
predicted <- predict(fit, data1)
However, I'm getting such good predictions that I have to assume that I am
not using these functions correctly and/or don't understand loess. One
explanation would be that predict(fit, data1) is actually using info from
data1 to refit the loess model.
Any clues would be appreciated as I try to figure this out.
[[alternative HTML version deleted]]
