# pcf.loclin: Multivariate local linear estimator In regpro: Nonparametric Regression

## Description

Computes the values of a multivariate local linear regression estimator on a regular grid.

## Usage

 `1` ```pcf.loclin(x, y, h, N, type=0, kernel="gauss", support=NULL, alt=FALSE, alt2=FALSE) ```

## Arguments

 `x` n*d data matrix; the matrix of the values of the explanatory variables `y` n vector; the values of the response variable `h` a positive real number; the smoothing parameter of the kernel estimate `N` vector of d positive integers; the number of grid points for each direction `type` integer 0,...,d; if type=0, then the regression function is estimated, otherwise the first partial derivative of the variable indicated by type is estimated `kernel` a character; determines the kernel function; either "gauss" or "uniform" `support` either NULL or a 2*d vector; the vector gives the d intervals of a rectangular support in the form c(low_1,upp_1,...,low_d,upp_d) `alt` an internal parameter `alt2` an internal parameter

## Value

a piecewise constant function

## Author(s)

Jussi Klemela

`pcf.kernesti`,
 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17``` ```set.seed(1) n<-100 d<-2 x<-8*matrix(runif(n*d),n,d)-3 C<-(2*pi)^(-d/2) phi<-function(x){ return( C*exp(-sum(x^2)/2) ) } D<-3; c1<-c(0,0); c2<-D*c(1,0); c3<-D*c(1/2,sqrt(3)/2) func<-function(x){phi(x-c1)+phi(x-c2)+phi(x-c3)} y<-matrix(0,n,1) for (i in 1:n) y[i]<-func(x[i,])+0.01*rnorm(1) num<-30 # number of grid points in one direction pcf<-pcf.loclin(x,y,h=0.5,N=c(num,num)) dp<-draw.pcf(pcf,minval=min(y)) persp(dp\$x,dp\$y,dp\$z,phi=30,theta=-30) contour(dp\$x,dp\$y,dp\$z,nlevels=30) ```