ppgmmga (version 1.3)

plot.ppgmmga: Plots the data onto the projection subspace estimated by the ppgmmga algorithm

Description

Plot method for objects of class 'ppgmmga'.

Usage

# S3 method for ppgmmga
plot(x, 
     class = NULL, 
     dim = seq(x$d), 
     drawAxis = TRUE, 
     bins = nclass.numpy,
     ...)

Value

Returns a object of class ggplot.

Arguments

x

An object of class 'ppgmmga' obtained from a call to ppgmmga function.

class

A numeric or character vector indicating the classification of the observations/cases to be plotted.

dim

A numeric vector indicating the dimensions to use for plotting. By default, all the dimensions of the projection subspace (i.e. x$d) are used. Subsets of all the available dimensions can also be provided (see example below.)
The resulting graph depends on the dimension: in the 1D case a histogram is provided, a scatterplot in the 2D case, a scatterplot matrix in higher dimensions.

drawAxis

A logical value specifying whether or not the axes should be included in the 2D scatterplot. By default is to TRUE.

bins

An R function to be used for computing the number of classes for the histogram. By default nclass.Sturges is used. Users may provide a different function. This argument only applies to 1D graphs.

...

further arguments.

Author

Serafini A.
Scrucca L.

Details

Plots the cloud of points onto a subspace after appling the Projection Pursuit algorithm based on Gaussian mixtures and Genetic algorithm implemented in ppgmmga function.

References

Scrucca, L. and Serafini, A. (2019) Projection pursuit based on Gaussian mixtures and evolutionary algorithms. Journal of Computational and Graphical Statistics, 28:4, 847–860. DOI: 10.1080/10618600.2019.1598871

See Also

ppgmmga

Examples

if (FALSE) {
data(iris)
X <- iris[,-5]
Class <- iris$Species

# 1D
pp1 <- ppgmmga(data = X, d = 1, approx = "UT")
summary(pp1, check = TRUE)
plot(pp1)
plot(pp1, Class)

# 2D
pp2 <- ppgmmga(data = X, d = 2, approx = "UT")
summary(pp2, check = TRUE)
plot(pp2)
plot(pp2, Class)

# 3D
pp3 <- ppgmmga(data = X, d = 3)
summary(pp3, check = TRUE)
plot(pp3)
plot(pp3, Class)
plot(pp3, Class, dim = c(1,3))
plot(pp3, Class, dim = c(2,3))
}