Let's assume that we have independent random variables and
and the
response variable
having the form
n = 100 t = normal(n,2) ; explanatory variable f1 = - sin(2*t[,1]) ; estimated functions f2 = t[,2]^2 eps = normal(n,1) * sqrt(0.75) ; error y = f1 + f2 +eps ; response variable
The data can come from praxis, too.
Our task is to estimate the unknown functions and
.
This chapter deals with such problems and their solutions. It ought to demonstrate and to explain how to use XploRe for nonparametric regression and data analysis in generalized additive models (GAM). It describes all quantlets which belong to the gam quantlib which contains all routines of XploRe provided for estimation and testing in generalized additive models. It also has several links to the gplm quantlib for generalized partial linear models (GPLM) in XploRe thus many quantlets which are used in gam are fully described in Chapter 6 but not mentioned here.