pairs: Pairs plot method for compositions

Description

Pairs plot function for compositions, allowing flexible representations.

Usage

# S3 method for acomp
pairs(x, labels, panel = vp.lrdensityplot, ...,
                         horInd = 1:ncol(x), verInd = 1:ncol(x),
                         lower.panel = panel, upper.panel = panel,
                         diag.panel = NULL, text.panel = textPanel,
                         label.pos = 0.5 + has.diag/3, line.main = 3,
                         cex.labels = NULL, font.labels = 1,
                         row1attop = TRUE, gap = 1, log = "")
 # S3 method for rcomp
pairs(x, labels, panel = vp.diffdensityplot, ...,
                         horInd = 1:ncol(x), verInd = 1:ncol(x),
                         lower.panel = panel, upper.panel = panel,
                         diag.panel = NULL, text.panel = textPanel,
                         label.pos = 0.5 + has.diag/3, line.main = 3,
                         cex.labels = NULL, font.labels = 1,
                         row1attop = TRUE, gap = 1, log = "")
  vp.lrdensityplot(x, y, col=2,..., alpha = NULL)
  vp.diffdensityplot(x, y, col=2,..., alpha = NULL)
  vp.lrboxplot(x, y, ...)
  vp.kde2dplot(x, y, grid=TRUE, legpos="bottomright", colpalette=heat.colors,...)

Arguments

x: a dataset of a compositional class; or for the panel functions, a vector of row components
y: for the panel functions, a vector of column components
...: further graphical parameters passed (see par)
labels: the names of the parts
panel: common panel function to use for all off-diagonal plots
horInd: indices of columns of x to plot on the horizontal axis, defaults to all columns
verInd: indices of columns of x to plot on the vertical axis, defaults to all columns
lower.panel: panel function for the lower triangle of plots, defaults to the common panel
upper.panel: panel function for the uppper triangle of plots, defaults to the common panel
diag.panel: panel function for the diagonal of plots, defaults to text.panel
text.panel: panel function to write labels on the diagonal panels
label.pos: y position of labels in the text panel
line.main: if main is specified, line.main gives the line argument to mtext() which draws the title. You may want to specify oma when changing line.main
cex.labels: graphics parameters for the text panel
font.labels: graphics parameters for the text panel
row1attop: logical. Should the layout be matrix-like with row 1 at the top, or graph-like with row 1 at the bottom?
gap: distance between subplots, in margin lines
log: a character string indicating if logarithmic axes are to be used: see plot.default. Should not be used and left to the panel function to handle
col: color for density and histogram components of the panel vp.*density
alpha: alpha level for marking normality in the panels vp.*density; default to no mark
grid: should a unit-grid be added to each panel?
legpos: where should the legend be placed? to be given as x in graphics::legend
colpalette: which color palette is desired for the 2d density levels?

Author

Raimon Tolosana-Delgado

Details

The data is displayed in a matrix of plots, after the indications of a panel function. This is a simple implementation of pairs compositional methods, the real functionality is controlled by the panel functions.

The three panel functions included here can be used for generating either boxplots or histograms plus kernel density plots of all pairwise logratios (in acomp) or differences (in rcomp) of the components. In the cas of histograms, these can be colored or left black-and-white depending on the adjustment to normality, controlled by a shapiro.test and the alpha-level given. These panel functions serve also as examples of how to generate user defined panels.

Examples

Run this code

data(SimulatedAmounts)
pairs(acomp(sa.lognormals))
pairs(rcomp(sa.lognormals))