Pollution models with multi-scale interactions
The meuse dataset contains some predictor variables that are on the same scale (x, y), and some that are on different scales (elev, dist, om). In a previous exercise, you fit a model where you predicted cadmium pollution as a function of location and elevation:
mod <- gam(cadmium ~ s(x, y) + s(elev),
data = meuse, method = "REML")
In this exercise, you'll build a model that allows multiple variables to interact despite these different scales using a tensor smooth, te().
This exercise is part of the course
Nonlinear Modeling with Generalized Additive Models (GAMs) in R
Exercise instructions
- Convert this to a model where
x,y, andelevall interact in a singlete()term, varying on their own scales. - Then summarize the model and visualize it with
plot().
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Fit the model
tensor_mod <- ___
# Summarize and plot
___
___