ggbiplot: Generic ggplot biplot function

In low-decarie/ggpca: Creates ggplot biplot methods for pca objects

Description

Produces a ggplot biplot from an PCA analysis similar to biplot. Currently has methods for princomp and prcomp.

Usage

ggbiplot(PC, selected.pc = c(1, 2), groups = NULL, scale = 1,
  length.alpha = T, ...)

Arguments

PC	a fitted object
selected.pc	vector of length 2 number of two principle components
groups	vector group assigned to each point
scale	scaling for explanatory variable arrows
length.alpha	logical wether transparency of explanatory variables should vary with length
...	other arguments to be passed to qplot
varnames

Details

If groups acts as stats:::biplot If groups is specified and has 6 or less unique values, observations are presented as different shapes for each group (ggplot shape argument) and circled with a 95 If groups is specified and has more than 6 unique values, observations are presented with different colours for each group (ggplot colour argument) and circled with a 95

Value

a ggplot object

See Also

biplot, prcomp, princomp

Examples

# Can be used on the results of princomp and prcomp
p <- ggbiplot(PC=princomp(iris[,-5]))
print(p)
# For 6 or less groups, uses shape and linetype
p <- ggbiplot(PC=prcomp(iris[,-5]),
                   selected.pc=c(1,2),
                   groups=iris[,5],
                   main="PCA of iris dataset")
print(p)
# For more than 6 groups, uses colour
p <- ggbiplot(PC=prcomp(iris[,-5]),
                   selected.pc=c(1,2),
                   groups=sample(LETTERS[1:7],nrow(iris), replace=T),
                   main="PCA of iris dataset")
print(p)

low-decarie/ggpca documentation built on May 21, 2019, 7:50 a.m.