FUNFITS data analysis and statistical tools for estimating functions

Abstract

A module of functions has been created to simplify and extend the fitting of curves and surfaces to data. Some important applications include estimates of the mean value of a measured response over some space of interest, response surface methods for designed experiments and the modeling of nonlinear dynamic systems. The core methods that are implemented in FUNFITS include prediction, plot and summary functions that help the user interpret results and visualize the estimate. FUNFITS is written mainly in the S programming language with some support from FORTRAN subroutines. All source is available and easy to modify. ·Contact Address: Department of Statistics Box 8203, North Carolina State University, Raleigh, NC, 27695-8203. Please visit the FUNFITS homepage to learn of revisions to this manual ( Go to the NCSU statistics department http://v'fIV .stat. ncsu. edu/ then to Douglas Nychka's home page under faculty. You will find FUNFITS as an item on this home page.) This work was supported in part by National Science Foundation Grant DMS-9217866-02, Becton Dickinson Research Center and the Environmental Protection Agency. We would also like to thank Jeffrey Andrew Royle, Qing Yang and Jerry Davis for their help on the spatial statistics functions.