vegan (version 2.7-2)

ordixyplot: Trellis (Lattice) Plots for Ordination

Description

Function ordixyplot provides an interface to plot ordination results using Trellis function xyplot in package lattice.

Usage

ordixyplot(x, data = NULL, formula, display = "sites", choices = 1:3,
    panel = "panel.ordi", aspect = "iso", envfit,
    type = c("p", "biplot"), ...)

Value

The function return Lattice objects of class

"trellis".

Arguments

x

An ordination result that scores knows: any ordination result in vegan and many others.

data

Optional data to amend ordination results. The ordination results are found from x, but you may give here data for other variables needed in plots. Typically these are environmental data.

formula

Formula to define the plots. A default formula will be used if this is omitted. The ordination axes must be called by the same names as in the ordination results (and these names vary among methods).

display

The kind of scores: an argument passed to scores.

choices

The axes selected: an argument passed to scores.

panel

The name of the panel function.

aspect

The aspect of the plot (passed to the lattice function).

envfit

Result of envfit function displayed in ordixyplot. Please note that this needs same choices as ordixyplot.

type

The type of plot. This knows the same alternatives as panel.xyplot. In addition ordixyplot has alternatives "biplot", "arrows" and "polygon". The first displays fitted vectors and factor centroids of envfit, or in constrained ordination, the biplot arrows and factor centroids if envfit is not given. The second (type = "arrows") is a trellis variant of ordiarrows and draws arrows by groups. The line parameters are controlled by trellis.par.set for superpose.line, and the user can set length, angle and ends parameters of panel.arrows. The last one (type = "polygon") draws a polygon enclosing all points in a panel over a polygon enclosing all points in the data. The overall polygon is controlled by Trellis parameters trellis.par.set plot.polygon and superpose.polygon.

...

Arguments passed to scores methods or lattice functions.

Author

Jari Oksanen

Details

The function provides an interface to the corresponding lattice function. All graphical parameters are passed to the lattice function so that these graphs are configurable. See Lattice and xyplot for details, usage and possibilities.

The argument x must always be an ordination result. The scores are extracted with vegan function scores so that these functions work with all vegan ordinations and many others.

The formula is used to define the models. Function has a simple default formula which is used if formula is missing. The formula must use the names of ordination scores and names of data.

The ordination scores are found from x, and data is optional. The data should contain other variables than ordination scores to be used in plots. Typically, they are environmental variables (typically factors) to define panels or plot symbols.

The proper work is done by the panel function. The layout can be changed by defining own panel functions. See panel.xyplot for details and survey of possibilities.

Ordination graphics should always be isometric: same scale should be used in all axes. This is controlled (and can be changed) with argument aspect in ordixyplot.

See Also

Lattice, xyplot.

Examples

data(dune, dune.env)
ord <- cca(dune)
## Scatter plot with polygons
ordixyplot(ord, data=dune.env, form = CA1 ~ CA2 | Management,
  groups=Manure, type = c("p","polygon"))
## Choose a different scaling
ordixyplot(ord, scaling = "sites")
## ... Slices of third axis
ordixyplot(ord, form = CA1 ~ CA2 | lattice::equal.count(CA3, 4),
   type = c("g","p", "polygon"))
## Display environmental variables
ordixyplot(ord, envfit = envfit(ord ~ Management + A1, dune.env, choices=1:3))