NW.fit {sharpData} | R Documentation |
This computes the Nadaraya-Watson local constant regression estimator at a single point. For multiple points, it is not efficient.
NW.fit(x, y, xgrid, h, kernel)
x |
numeric explanatory vector |
y |
numeric response vector |
xgrid |
numeric vector |
h |
numeric bandwidth |
kernel |
character constant |
a vector of predictor values
W.J. Braun
Wand and Jones, 1996, Kernel Smoothing, Chapman and Hall.
x <- sort(runif(100)) y <- x + sin(x) + rnorm(100)*.1 NW.fit(x,y,c(.3, .5, .7),h=.1)