[R] invalid type (list) for variable

[Jen Maresh]

Hello -

I am attempting to add interaction terms to a model using the gamm
function in the mgcv package. I stripped it down to only two terms to
simplify it. I have no trouble with individual terms, continuous or
factors, nor with interacting factor terms; I cannot, however, figure
out how to add interactions between continuous*factor terms. I applied
a smoother to my continuous term because it has a non-linear trend
over time. Below is an example of my attempt, followed by the error
message:

MMF.Test <- gamm(MaxDepth ~
            s(TripTime) + Trtmt +
            s(TripTime):Trtmt,
            random = list(Seal =~1), method = "REML",
            data = EsealsAll)


"Error in model.frame.default(formula = MaxDepth ~ Trtmt +
s(TripTime):Trtmt +  :
  invalid type (list) for variable 's(TripTime)'
In addition: Warning message:
In sp[i] <- ind :
  number of items to replace is not a multiple of replacement length"

I thought this was a problem with trying to use different vector
lengths, but the length of both is the same (1001). Help please? What
am I doing wrong? Not sure if my mistake is one of faulty statistical
reasoning, or incompetence with R.

Thank you in advance for any suggestions!


~~~~~~~~~~~~~~~~~~~~~
Jennifer Maresh, PhD Student
Center for Ocean Health, Long Marine Lab
100 Shaffer Rd.
University of California
Santa Cruz, CA 95060
maresh at biology.ucsc.edu