sieveplot: Sieve Plots

Description

Visualization of two-way contingency tables: plots rectangles with areas proportional to the expected cell frequencies and filled with a number of squares equal to the observed frequencies. Thus, the densities visualize the deviations of the observed from the expected values.

Usage

## S3 method for class 'default':
sieveplot(x, reverse_y = TRUE, type = c("observed", "expected"),
  main = deparse(substitute(x)),
  values = c("none", "cells", "margins", "both"),
  frequencies = c("absolute", "relative"),
  sieve_colors = c("red","blue"),
  sieve_lty = c("longdash", "solid"), exp_color = "gray",
  exp_lty = "dotted", margin = 0.01, newpage = TRUE, pop = TRUE,
  margins = c(4,3,4,4), xlab = names(dimnames(x))[2],
  ylab = names(dimnames(x))[1], ...)
## S3 method for class 'formula':
sieveplot(formula, data = NULL, \dots, subset)

Arguments

a two-way contingency table, as generated by table.

reverse_y

logical; if TRUE, the y axis is reversed (i.e., the rectangles' positions correspond to the contingency table).

type

switch indicating whether rectangles should be filled according to observed or expected frequencies.

main

character string specifying the title.

values

switch indicating whether the frequencies of cells or margins or of both should be plotted.

frequencies

switch indicating the type of these frequencies: relative or absolute.

sieve_colors, sieve_lty

vectors with up to two color/line type entries: the first is used for negative and the second for positive deviations from the expected frequencies.

exp_color, exp_lty

color/line type entry for the expected values grid.

margin

lines of margin between the cell rectangles.

margins

vector of margins (see par).

newpage

logical; if TRUE, the plot is drawn on a new page.

pop

logical; if TRUE, all newly generated viewports are popped after plotting.

xlab, ylab

labels of x- and y-axis.

formula

a formula, such as y ~ x. For details, see xtabs.

data

a data frame (or list), or a contingency table from which the variables in formula should be taken.

subset

an optional vector specifying a subset of the rows in the data frame to be used for plotting.

...

further graphics parameters (see gpar).

encoding

latin1

References

H. Riedwyl & M. Sch�pbach (1994), Parquet diagram to plot contingency tables. In F. Faulbaum (ed.), Softstat '93: Advances in Statistical Software, 293--299. Gustav Fischer, New York.

M. Friendly (2000), Visualizing Categorical Data, SAS Institute, Cary, NC.

Examples

Run this code

data(HairEyeColor)

## aggregate over 'sex':
(tab <- margin.table(HairEyeColor, c(2,1)))

## plot expected values:
sieveplot(tab, type = "expected", values = "both")

## plot sieve diagram:
sieveplot(tab)

## an example for the formula interface:
data(VisualAcuity)
sieveplot(Freq ~ right + left,  data = VisualAcuity, reverse_y = FALSE)