g.princals

This script provides a way to enhance the appearance of plots created with the princals function from the Gifi package in R. Since princals generates complex outputs that require additional processing for effective visualization, the script uses ggplot2 to improve the aesthetic and clarity of the results.

Contents

• Proposal
• Instruction
• R code
• Output
• Reference

Proposal

The g.princals function is specifically designed to improve the aesthetics of plots generated by the plot.princals function from the Gifi package, which by default relies on the base R graphics package for visualizations. By harnessing the powerful capabilities of ggplot2, a package renowned for producing high-quality, customizable graphics, g.princals offers a more polished and visually appealing experience. This function provides options to create various plots, including transformed plots, scree plots, loading plots, and biplots for hierarchical categorical variables.

Instruction

"g.princals" replaces the default "plot" function in R, generating customized graphics for customer items.

• x: call a object was generated by princals(dataset).
• plot.dim: Vector with dimensions to be plotted with the default condition being set to c(1, 2).
• var.subset: with the default condition being set to"all"
• max.plot.array: with the default condition being set to 2.
• stepvec: Vector whether step function should be plotted (see details in princals/R Documentation).with the default condition being set to NA.
• col.lines: colour of line. Default condition being set to "black".
• main: Title of picture. Default condition being set to name of princals object.
• show: Options to choose are TRUE,FALSE,T,F,1 and 0. Default condition being set to "1" mean it shows plots.
• save: Options to choose are TRUE,FALSE,T,F,1 and 0. Default condition being set to "1" mean it saves plots.
• name: label or name of princals object. Default condition being set to name of princals object.
• units: unit of width/height in 'mm'
• width: Default condition being set to 250mm.
• height: Default condition being set to 250mm.
• res: resolution of pictures. Default condition being set to 300 pixels.
• dispersion: dispersion of scores of object in loadplot.it is decreased overdispersion or underdispersion. Default condition being set to 3, it means, mean±3*standard deviation.
• point.size.loadplot: it is point size of loadplot. Default condition being set to 1.
• point.alpha.loadplot: it is point alpha(transparency) of loadplot. Default condition being set to 0.1,
• language: choose a English or Spanish like language. Default condition being set English.
• legend.position: Options to choose are none, left, right, bottom, and top. Default condition being set to right. it means legend position is to right.
• legend.group: Options to choose are TRUE,FALSE,T,F,1 and 0. Default condition being set to "1" mean it shows legend.

R code

Finally, it were showed a picture which generated by "g.princals". If the label names are lengthy, it's advisable to shorten them into acronyms.

################################################################################
setwd("~/file")
source("https://github.com/jasb3110/g.princals/blob/f286b26f79da57600c3e7f207a6f653f24028a76/g.princals.R)

library("Gifi")
# ordinal PCA
fitord <- princals(ABC,ndim = 2)## ordinal PCA
g.princals(fitord,save=0,show=1,dispersion =3,legend.group = 0)#show plots
g.princals(fitord,save=0,show=1,dispersion =3,colour.group = c(rep("A",5),rep("B",6)),legend.group = 1,language = "spanish")#show plots

g.princals(fitord,save=1,show=1,dispersion =3,legend.group = 0)#save plots in PNG format
g.princals(fitord,save=1,show=0,dispersion =3,colour.group = c(rep("A",5),rep("B",6)),legend.group = 1,language = "spanish")#save plots in PNG format
################################################################################

Output

In this section, an R script and corresponding plots created using the g.princals function are presented in Spanish, and group of variables.

Picture. Transplots of ABC dataset, just 1 to 4 variables

Picture. Transplots of ABC dataset, just 5 to 8 variables

Picture. Transplots of ABC dataset, just 9 to 11 variables

Picture. Screeplot of ABC dataset

Picture. Loadplot of ABC dataset

Picture. Biplot of ABC dataset

Reference

• https://rdrr.io/rforge/Gifi/src/R/plot.princals.R
• https://github.com/jasb3110/g.princals/blob/0ed7b04adcb1039de5f9cafca0c6315a36a37e09/g.princals.R