Preface to First Edition - lib

ANALYSIS USING R 189R> x y men1500m_lowess plot(time ~ year, data = men1500m1900)R> lines(men1500m_lowess, lty = 2)R> men1500m_cubic lines(x, predict(men1500m_cubic), lty = 3)time215 220 225 230 235 240 2451900 1920 1940 1960 1980 2000yearFigure 10.4Scatterplot of year and winning time with fitted values from a smoothnon-parametric model.10.3.2 Air Pollution in US CitiesUnfortunately, we cannot fit an additive model for describing the SO 2 concentrationbased on all six covariates because this leads to more parametersthan cities, i.e., more parameters than observations when using the defaultparameterisation of mgcv. Thus, before we can apply the gam function frompackage mgcv, we have to decide which covariates should enter the model andwhich subset of these covariates should be allowed to deviate from a linearregression relationship.As briefly discussed in Section 10.2.3, we can fit an additive model using theiterative boosting algorithm as described by Bühlmann and Hothorn (2007).© 2010 by Taylor and Francis Group, LLC