k.nn(haerdle) | R Documentation |
k-nearest neighbour non-parametric regression
k.nn(x, y, k)
x |
data vector |
y |
data vector |
k |
number of neighbours |
list of components m (the estimate at the points of x), x and y
`Smoothing Techniques with Implementation in S', Wolfgang Haerdle, Springer, 1991
data(faithful) # Figure 5.6 par(mfrow=c(2,2)) plot(faithful$eruptions,faithful$waiting) f.knn<-k.nn(faithful$eruptions,faithful$waiting,5) lines(f.knn$x,f.knn$m) plot(faithful$eruptions,faithful$waiting) f.knn<-k.nn(faithful$eruptions,faithful$waiting,10) lines(f.knn$x,f.knn$m) plot(faithful$eruptions,faithful$waiting) f.knn<-k.nn(faithful$eruptions,faithful$waiting,20) lines(f.knn$x,f.knn$m) plot(faithful$eruptions,faithful$waiting) f.knn<-k.nn(faithful$eruptions,faithful$waiting,30) lines(f.knn$x,f.knn$m)