plot.kepdf: Plot objects of class kepdf

In pdfCluster: Cluster Analysis via Nonparametric Density Estimation

Plot objects of class kepdf

Description

Functions and methods for plotting objects of kepdf-class.

Usage

## S4 method for signature 'kepdf'
plot(x, y, eval.points = NULL, n.grid = NULL,
		data = NULL, add = FALSE, main = NULL, xlab = NULL, ylab = NULL, 
		zlab = NULL, col = NULL, col.data=2, type="l", props = c(75,50,25), 
		method="contour", ticktype = "detailed", indcol = NULL, 
		text.diag.panel = NULL, gap = 0.5, ...)

Arguments

x	An object of kepdf-class;
y	Not used; for compatibility with generic plot;
eval.points	A matrix of data points at which the density to be plotted has to be evaluated; the number of columns must correspond to the dimension of the sample data used to estimate the density. If not provided, density is evaluated on a grid defined on the range of sample data.
n.grid	A vector with length set to the number of column of sample data, defining the length of the grid on which the density to be plotted is evaluated; this argument is ignored when eval.points is not NULL.
data	Data to be optionally superimposed to the density plot.
add	Logical. If TRUE, add to a current plot.
main	An overall title for the plot.
xlab	A title for the x axis.
ylab	A title for the y axis.
zlab	A title for the z axis.
col	A specification for the plotting color.
col.data	A specification for the color of data. Ignored if data is NULL.
type	What type of plot should be drawn. This argument applies when kernel density estimate is performed on unidimensional data only. Default value is "l".
props	A vector defining the fraction of the data to be included within each density level. This argument applies when kernel density estimate is performed on multidimensional data only.
method	One of c("contour", "image", "perspective"). To be used when two or higher dimensional data have been used to estimate the density.
ticktype	Character: "simple" draws just an arrow parallel to the axis to indicate direction of increase; "detailed" draws normal ticks; to be used if method="perspective" only.
indcol	Vector of the column positions to be plotted, when densities are estimated on higher than 2-dimensional data.
text.diag.panel	Text to be displayed on the diagonal panels when plotting densities estimated on higher than 2-dimensional data.
gap	Distance between subplots, when plotting densities estimated of 2-dimensional data or higher-dimensional data.
...	Further arguments to be passed to plot, image, contour, persp.

Details

When density estimation is based on two or higher dimensional data, these functions make use of functions contour, image and persp. For densities estimated on higher than 2-d data, the pairwise marginal estimated densities are plotted for all possible pairs of coordinates or a chosen selection of them.

Value

A list containing the following elements:

eval.points	data points at which the plotted density has been evaluated
estimate	the estimated density at eval.points

Methods

signature(x = "kepdf", y = "missing")

S4 method for plotting objects of kepdf-class.

See Also

kepdf-class, plot, contour, image, plot-methods,persp

Examples

#1-d example
set.seed(123)
x1 <- rnorm(50)

#normal optimal bandwidth
pdf1a <- kepdf(x1)
#shrink the smoothing parameter
pdf1b <- kepdf(x1, h=0.5*h.norm(x1))

plot(pdf1a, n.grid=50, data=x1, xlab="x1", ylim=c(0, max(c(pdf1a@estimate,
   pdf1b@estimate))))
plot(pdf1b, n.grid=50, lty=2, add=TRUE)

#2-d example
set.seed(123)
x2 <- cbind(rnorm(50),rnorm(50))

pdf2 <- kepdf(x2)

plot(pdf2, n.grid=c(50,50), data=x2)
plot(pdf2, n.grid=c(50,50), method="image")
plot(pdf2, n.grid=c(50,50), method="perspective", phi=30, theta=30)

#3-d example
set.seed(123)
x3 <- cbind(rnorm(50), rnorm(50), rnorm(50))

pdf3 <- kepdf(x3)

plot(pdf3, n.grid=c(50,50,50))
plot(pdf3, n.grid=c(50,50,50), method="image", col = terrain.colors(30))
plot(pdf3, n.grid=c(50,50,50), method="perspective", phi=30, theta=30)

pdfCluster documentation built on Dec. 2, 2022, 5:14 p.m.