NW.Warping(haerdle)R Documentation

Nadaraya-Watson non-parametric regression via WARPing

Description

Nadaraya-Watson non-parametric regression via WARPing

Usage

NW.Warping(x, y, h, M=10, kernel=4, na.handling=0)

Arguments

Required:
x data vector
y data vector
h bandwidth
M number of small bins
kernel code for kernel. 1 = uniform, 2 = triangle (ASH), 3 = Epanchenikov, 4 = quartic, 5 = triweight.
na.handling control handling of 0/0

Value

list with components midpoints, m (fitted curve), x and y. The curve is evaluated at the midpoints of the(small) bins

References

`Smoothing Techniques with Implementation in S', Wolfgang Haerdle, Springer, 1991

Examples

data(dat.reg)
plot(dat.reg)
lines(dat.reg$x, dat.reg$m)
nw<-NW.Warping(dat.reg$x, dat.reg$y, 0.05)
lines(nw$midpoints, nw$m, lty=2)
nw<-NW.Warping(dat.reg$x, dat.reg$y, 0.10)
lines(nw$midpoints, nw$m, lty=3)


[Package Contents]