Projection Pursuit Based on Gaussian Mixtures and Evolutionary Algorithms

Projection Pursuit (PP) algorithm for dimension reduction based on Gaussian Mixture Models (GMMs) for density estimation using Genetic Algorithms (GAs) to maximise an approximated negentropy index. For more details see Scrucca and Serafini (2019) .

An R package accompanying the paper Projection pursuit based on Gaussian mixtures and evolutionary algorithms by Luca Scrucca and Alessio Serafini (2018).

Installation

You can install the released version of ppgmmga from CRAN:

install.packages("ppgmmga")

or the development version from GitHub:

# install.packages("devtools")
devtools::install_github("luca-scr/ppgmmga")

Usage

Usage of the main functions and several examples are included in the papers shown in the references section below.

For an intro see the vignette A quick tour of ppgmmga, which is available as

vignette("ppgmmga")

Note that if the package is installed from GitHub the vignette is not automatically created. However, it can be created when installing from GitHub with the code:

devtools::install_github("luca-scr/ppgmmga", build_vignettes = TRUE)

References

Scrucca, L. and Serafini, A. (2018) Projection pursuit based on Gaussian mixtures and evolutionary algorithms. Under review.