ggbiplot: methods for making biplots from various projection &...

In zdk123/compPLS: Conduct Partial Least Squares regression for compositional data

methods for making biplots from various projection & classification models

Description

Pretty biplots using ggplots

Usage

ggbiplot(xobj, ...)

## S3 method for class 'princomp'
ggbiplot(xobj, ...)

## S3 method for class 'prcomp'
ggbiplot(xobj, ...)

## S3 method for class 'lda'
ggbiplot(xobj, ...)

## S3 method for class 'plsda'
ggbiplot(xobj, Yplot = FALSE, ...)

## S3 method for class 'splsda'
ggbiplot(xobj, ...)

## S3 method for class 'matrix'
ggbiplot(xobj, ...)

## Default S3 method:
ggbiplot(
  xobj,
  grouping,
  select = 1:2,
  circle = FALSE,
  circle.prob = 0.69,
  plot.loadings = TRUE,
  label.loadings = FALSE,
  sub.loadings = 1:nrow(xobj$loadings),
  label.offset = 0,
  label.size = 4.5,
  scale.loadings = 1,
  col.loadings = scales::muted("red"),
  alpha = 1,
  col = grouping,
  shape = NULL,
  group.ellipse = FALSE,
  scale.ellipse = 1,
  group.cloud = FALSE,
  xlab = "",
  ylab = "",
  equalcoord = TRUE,
  size = 3,
  size.loadings = 1,
  loadingsOnTop = FALSE
)

Arguments

xobj	The object to be plotted
grouping	an optional grouping vector (ie - for coloring points)
select	index of components to be plotted (must be length 2)
circle	enclose points in a circle
circle.prob	controls circle diameter (scales data std dev) if circle = TRUE
plot.loadings	should loading vectors be plotted
label.loadings	text of loadings labels, taken from rownames of loadings (depends on class of xobj)
label.offset	absolute offset for loading labels, so labels don't cover loadings vectors
scale.loadings	scale length of loading vectors for plotting purposes
col.loadings	a single value of vector for color of loadings
alpha	controls relative transparency of various plot features
col	color factor for points
group.ellipse	enclose within-group points in an covariance ellipse
scale.ellipse	scale group.ellipse to 1 standard deviation
group.cloud	connect within-group points to a group mean point with a straight edge
xlab	label for x axis
ylab	label for y axis
equalcoord	equal coordinates, ie should the plot area be square?
size	point size
size.loadings	line width of loading vectors

Details

additional plotting attributes (eg colors, themes, etc) can be chained on in the usual way for ggplots

Examples

 # an LDA example with iris data
ldamod <- lda(iris[,1:4], grouping=iris[,5])
ggbiplot(ldamod, grouping=iris[,5], alpha=.7, group.cloud=TRUE) + theme_bw()

zdk123/compPLS documentation built on April 24, 2022, 2:44 p.m.