NonpModelCheck (version 2.0)

plot3d.localpoly.reg: 3d plot from a local polynomial fit

Description

Create a 3d plot from a local polynomial fit of two covariates and a response variable.

Usage

plot3d.localpoly.reg(X,Y, bandwidth = "CV", gridsize = 30,  degree.pol = 0, kernel.type = "epanech", gridsurface = 30,  xlab=expression(X_1), ylab=expression(X_2),  zlab=expression(Y), theta = 30, phi = 30, expand = 0.5,  col = "lightblue", ltheta = 120, shade = 0.75,  ticktype = "detailed", pch = 16,...)

Arguments

X

n by 2 matrix with observations, rows corresponding to data points and columns correspond to covariates.

Y

vector of observed responses.

bandwidth

bandwidth, vector or matrix. If "CV", leave-one-out cross validation with criterion of minimum MSE is performed to select a unique bandwidth that will be used for all dimensions of X; if "GCV", Generalized Cross Validation is performed to select a unique bandwidth that will be used for all dimensions of X; if "CV2" leave-one-out cross validation for each covariate; and if "GCV2", GCV for each covariate. It may be a vector for each dimension of the X; or a matrix of bandwidths (not to be confused with bandwidth matrix H), where each row is a vector of size 2, representing a bandwidth that changes with the location of estimation for the grid. See localpoly.reg.

gridsize

number of possible bandwidths to be searched in cross-validation. If left as default 0, gridsize is taken to be 5+as.integer(100/d^3). If cross-validation is not performed, it is ignored.

degree.pol

degree of the polynomial to be used in the local fit.

kernel.type

kernel type, options are "box", "trun.normal", "gaussian", "epanech", "biweight", "triweight" and "triangular". "trun.normal" is a gaussian kernel truncated between -3 and 3.

gridsurface

number of points on each axis at which to estimate the local polynomial surface.

xlab

parameter for persp

ylab

parameter for persp

zlab

parameter for persp

theta

parameter for persp

phi

parameter for persp

expand

parameter for persp

col

parameter for persp

ltheta

parameter for persp

shade

parameter for persp

ticktype

parameter for persp

pch

parameter for persp

...

further parameters for plotting persp

Value

X

the same input matrix

Y

the same input response vector

points

points at which to get smoothed values

bandwidth

bandwidth used for the polynomial fit

predicted

matrix with the predicted values at grid points

Details

Uses function "persp" to plot the estimated surface of a local polynomial fit in a nonparametric model with two covariates. The surface is estimated at points of a grid with size "gridsurface", which are evenly distributed between the minimum and maximum of the observed predictors. It also adds the observed points to the plot.

See Also

localpoly.reg

Examples

X = matrix(0,50,2)
X[,1] = runif(50,-2,2)
X[,2] = runif(50,-2,2)
Y = 4*sin(pi*X[,1]) + X[,2] + rnorm(50)

plot3d.localpoly.reg(X,Y, bandwidth="CV2", gridsize = 15, 
degree.pol = 0, gridsurface=20)