Due Date: Friday, February 10, 2012
The file islands.txt contains the data appearing in McMaster (2005). This is a comma-delimited text file in which the variable names appear in the first row.
McMaster (2005) examined native and non-native plant species richness on 22 islands off the coast of the northeastern United States and maritime Canada and studied its relationship to various geographic factors. For this exercise we will focus on only two variables.
Fit the following five models to the variable species richness S as a function of island area A.
In all models β0 and β1 are parameters to be estimated. In the normal models an additional parameter σ2 needs to be estimated, while the negative binomial model has the dispersion parameter θ to be estimated. The Poisson model has no additional parameters. Observe that models 1 and 3 can be fit as ordinary regression models (general linear models), model 2 is a nonlinear model (fit using nonlinear least squares), and models 4 and 5 are generalized linear models proper.
For each model,
The second model is nonlinear and can be fit in R using the nls function. Check the help page for nls to see how to use it. Basically in addition to the nonlinear formula you need to supply starting values for the parameters in the form of a list. The list is then entered as the start argument of nls. Observe that in creating the list you need to specify the starting values by their names. Fit the log-Arrhenius model first (model 3) and exponentiate the parameter estimates you obtain to use as starting estimates in this model, model 2.
Jack Weiss Phone: (919) 962-5930 E-Mail: jack_weiss@unc.edu Address: Curriculum for the Environment and Ecology, Box 3275, University of North Carolina, Chapel Hill, 27599 Copyright © 2012 Last Revised--February 3, 2012 URL: https://sakai.unc.edu/access/content/group/2842013b-58f5-4453-aa8d-3e01bacbfc3d/public/Ecol562_Spring2012/docs/assignments/assign4.htm |