ecdfHT (version 0.1.1)

ecdfHT.draw: Graph and annotate an ecdfHT plot

Description

Does the computations and plotting for ecdfHT and can be used to add to an existing plot.

Usage

ecdfHT.draw(transform.info, x, p, show.plot = TRUE, new.plot = FALSE, show.ci = FALSE, xlab = "x", ylab = "", ...)
ecdfHT.axes(transform.info, x.labels = c(), y.labels = c(), show.vert.gridlines = FALSE, show.horiz.gridlines = FALSE, ...)
ecdfHT.h(x, t)
ecdfHT.g(p, q)

Arguments

transform.info

A list with information about the transformation, computed in ecdfHT

x

The data, a vector of double precision numbers. Assumbed to be sorted and have distinct values.

p

Probabilities, a vector of doubles. Typically p[i]=(i=0.5)/length(x), unless there are repeats in x.

show.plot

Boolean value: indicates whether to plot or not.

new.plot

Boolean value: indicates whether to produce a new plot or add to an existing plot.

show.ci

Boolean value: indicates whether or not confidence intervals are shown.

xlab

String to label the horizontal axis.

ylab

String to label the vertical axis.

...

Optional parameters for the plot, e.g. col='red'.

x.labels

Vector of numbers specifying the location of the labels on the horizontal axis

y.labels

Vector of numbers specifying the location of the labels on the vertical axis

show.vert.gridlines

Boolean value indicating whether or not vertical grid lines should be drawn.

show.horiz.gridlines

Boolean value indicating whether or not horizontal grid lines should be drawn.

t

A vector of length 3 that specifies the x values that determine the left tail, middle, and right tail

q

A vector of length 3 that specifies the quantile values that determine the left tail, middle, and right tail.

Value

A list of values used in the plot, see return value of ecdfHT.ecdfHT.h returns the vector y=h(x;t), ecdfHT.g returns the vector y=g(p;q)

Details

ecdfHT.draw computes transform and plots. ecdfHT.axes draws axes on the plot; it can be used to manually select tick marks, etc. ecdfHT.h computes the function h(x) for the transformation of the horizontal axis. ecdfHT.g computes the function g(p) for the transformation of the vertical axis.

Always call ecdfHT first to produce the basic plot, then use ecdfHT.draw to add other curves to the plot as in the examples below

Examples

set.seed(1)
x <- rcauchy( 1000 )
t.info <- ecdfHT( x, show.axes=FALSE )
ecdfHT.axes( t.info, x.labels=c(-50,-5,0,5,50), y.labels=c(.001,.01,.1,.5,.9,.99,.999),
  show.vert.gridlines=TRUE, show.horiz.gridline=TRUE, lty=2 )
q1 <- qcauchy(t.info$ecdf) # Cauchy quantiles
ecdfHT.draw( t.info, q1, t.info$ecdf, col='red',show.ci=TRUE)
q2 <- qnorm(t.info$ecdf,sd=sd(x))  # Gaussian quantiles
ecdfHT.draw( t.info, q2, t.info$ecdf, col='green',show.ci=TRUE)
title(paste("simulated Cauchy data, n=",length(x),"\nred=Cauchy cdf, green=normal cdf"))

x <- seq(-5,5,1)
t <- c(-3,0,3)
ecdfHT.h(x,t)
p <- seq(0.05,.95,.1)
q <- c(.1,.5,.9)
ecdfHT.g(p,q)