simpleSmoothers: Simple smoother

In locpol: Kernel Local Polynomial Regression

Simple smoother

Description

Computes simple kernel smoothing

Usage

  simpleSmootherC(x, y, xeval, bw, kernel, weig = rep(1, length(y)))
  simpleSqSmootherC(x, y, xeval, bw, kernel)

Arguments

x	x covariate data values.
y	y response data values.
xeval	Vector with evaluation points.
bw	Smoothing parameter, bandwidth.
kernel	Kernel used to perform the estimation, see Kernels
weig	weights if they are required.

Details

Computes simple smoothing, that is to say: it averages y values times kernel evaluated on x values. simpleSqSmootherC does the average with the square of such values.

Value

Both functions returns a data.frame with

x	x evaluation points.
reg	the smoothed values at x points.

...

Author(s)

Jorge Luis Ojeda Cabrera.

See Also

PRDenEstC, Kernel characteristics

Examples

	size <- 1000
	x <- runif(100)
	bw <- 0.125
	kernel <- EpaK
	xeval <- 1:9/10
	y <- rep(1,100)	
	##	x kern. aver. should give density f(x)
	prDen <- PRDenEstC(x,xeval,bw,kernel)$den
	ssDen <- simpleSmootherC(x,y,xeval,bw,kernel)$reg
	all(abs(prDen-ssDen)<1e-15)
	##	x kern. aver. should be f(x)*R2(K) aprox.
	s2Den <- simpleSqSmootherC(x,y,xeval,bw,kernel)$reg
	summary( abs(prDen*RK(kernel)-s2Den) )
	summary( abs(1*RK(kernel)-s2Den) )
	##	x kern. aver. should be f(x)*R2(K) aprox.
	for(n in c(1000,1e4,1e5))
	{
		s2D <- simpleSqSmootherC(runif(n),rep(1,n),xeval,bw,kernel)$reg
		cat("\n",n,"\n")
		print( summary( abs(1*RK(kernel)-s2D) ) )
	}

locpol documentation built on Nov. 29, 2022, 9:05 a.m.