dependenceAnalysisPlot: Bivariate dependence analysis plot

In mathphysmx/empiricalDistribution: Empirical Distributions

Description

Plot scatterplots and pseudo-observations plots over a rectangular domain partition by rectangular areas (also called pixes). This function can be used with the function plottingPosition for the pseudo-observations.

Usage

dependenceAnalysisPlot(x = NULL, asp = 1, main = "", xGridBreaks = c(0,
  0.25, 0.5, 0.75, 1), yGridBreaks = xGridBreaks, xlab = "x", ylab = "y",
  xlabels = TRUE, ylabels = TRUE, xlSide = 1, ylSide = 2,
  col = par("col"), colp = c("white", "gray90"), grid = NULL, ...)

Arguments

x	table (2-colum matrix or data.frame) of, possibly pseudo-, observations
asp	The y/x aspect ratio, see plot.window.
main	Character. Main title for the plot. See title.
xGridBreaks, yGridBreaks	zone delimiters. Default to Inter-Quartile Area (iqa := c(0, .25, .5, .75, 1) x c(0, .25, .5, .75, 1)).
xlab, ylab	Character. Label of the axis.
xlabels, ylabels	Character. Labels of the annotations of the axis ticks marks.
xlSide, ylSide	Side where the labels of the axis must be plotted. See side parameter of axis.
col	color of the points.
colp	color of the pixels.
grid	list. grid constructed with (posibly pseudo-) observations of modes, minima, inflexion points or any other points of interest of both variables. grid is passed to abline.
...	further parameters to points.default.

Author(s)

Francisco Mendoza-Torres ([e-mail])

Examples

# Example 1: pseudo-observations
ne <- 30
ue <- matrix(runif(ne), ne/2, 2)
xpixelse <- c(0, .25, .5, .75, 1)
dependenceAnalysisPlot(x = ue, xGridBreaks = xpixelse) # colp = rainbow(n = npixels),

# Example 2: observations
ne <- 100
ue <- matrix(rnorm(ne), ne/2, 2)
xpixelse <- quantile(ue[, 1])
ypixelse <- quantile(ue[, 2])
dependenceAnalysisPlot(x = ue,  asp = NA,
    xlabels = c("min", "q1", "M", "q3", "max"),
    xGridBreaks = xpixelse,
    yGridBreaks = ypixelse)

npixels <- (length(xpixelse) - 1) * (length(ypixelse) - 1)
dependenceAnalysisPlot(x = ue,
    xGridBreaks = xpixelse,
    yGridBreaks = ypixelse,
    colp = rainbow(n = npixels),
    pch = 20, cex = 2, asp = NA,
    grid = list(h = 0, v = c(-1, 1),
                lty = 2, col = "blue"))

mathphysmx/empiricalDistribution documentation built on May 7, 2019, 10:55 p.m.