prinsimp (version 0.8-8)

plot.simpart: Plot method for Simple Partition

Description

Produces plots for the model basis vectors and the vectors forming the simplicity basis for the nearly null space, along with the variance-simplicity view and the percent of variance explained panel. Only six basis vectors can be displayed at a time.

Usage

"plot"(x, display = list(model=TRUE, simple=TRUE), layout, ...)

Arguments

x

an object of class "simpart", typically result of simpart.

display

a list. Specifies which model and simplicity basis vectors to plot. Must be consistent with x.

layout

a matrix of locations for the component plots, following the format used in the layout function. The default is to draw basis plots in the left half of the figure area, arranged in a U-curve with model basis starting at the top of the left arm and simple basis at the top right. The variance-simplicity view and percent of variance explained panel take up the right half of the figure.

The subplots are drawn in the following order: first model then simple basis, followed by the variance-simplicity plot, and variance-explained last.

...

other parameters passed through to plotting functions

Details

The variance-simplicity view and the percent of variance explained panel are always produced everytime plot.simpart is called. The variance-simplicity view plots the percent of variance explained and the simplicity values of the basis vectors specified in display. If the number of basis vectors is greater than 6 and the user specifies more than 6 in display, an error message is generated. In display, non-existent basis numbers are ignored.

See Also

basisplot, varsimp, varperc, simpart

Examples

library(prinsimp)
require(graphics)

## Caterpillar estimated covariance matrix from Kingsolver et al (2004)
## Measurements are at temperatures 11, 17, 23, 29, 35, 40
data(caterpillar)

## Default plot method for a 2-dimensional nearly null space
cat.sim <- simpart(caterpillar, simpledim = 2,
                   x = c(11, 17, 23, 29, 35, 40), cov = TRUE)
plot(cat.sim)

## Plot only the two simplicity basis vectors in the 2-dimensional
## nearly null space
plot(cat.sim, display = list(simple=1:2))
    
## Try, by mistake to plot first four simple vectors when simpledim=2.
## Will plot only 2.
plot(cat.sim, display = list(simple=1:4))

## Only plot the first two model basis vectors, the variance simplicity
## view, and the percent variance explained panel using the 'layout'
## argument
plot(cat.sim, display = list(model=1:2), layout = matrix(1:4, nrow=2, ncol=2))