fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining


Description

Draw the graph of individuals/variables from the output of Principal Component Analysis (PCA).

The following functions, from factoextra package are use:

  • fviz_pca_ind(): Graph of individuals
  • fviz_pca_var(): Graph of variables
  • fviz_pca_biplot() (or fviz_pca()): Biplot of individuals and variables

Install and load factoextra

The package devtools is required for the installation as factoextra is hosted on github.

# install.packages("devtools")
library("devtools")
install_github("kassambara/factoextra")

Load factoextra :

library("factoextra")

Usage

# Graph of individuals
fviz_pca_ind(X, axes = c(1, 2), geom = c("point", "text"),
       label = "all", invisible = "none", labelsize = 4,
       pointsize = 2, habillage = "none",
       addEllipses = FALSE, ellipse.level = 0.95, 
       col.ind = "black", col.ind.sup = "blue", alpha.ind = 1,
       select.ind = list(name = NULL, cos2 = NULL, contrib = NULL),
       jitter = list(what = "label", width = NULL, height = NULL),  ...)
# Graph of variables
fviz_pca_var(X, axes = c(1, 2), geom = c("arrow", "text"),
       label = "all", invisible = "none", labelsize = 4,
       col.var = "black", alpha.var = 1, col.quanti.sup = "blue",
       col.circle = "grey70",
       select.var = list(name =NULL, cos2 = NULL, contrib = NULL),
       jitter = list(what = "label", width = NULL, height = NULL))
# Biplot of individuals and variables
fviz_pca_biplot(X, axes = c(1, 2), geom = c("point", "text"),
   label = "all", invisible = "none", labelsize = 4, pointsize = 2,
    habillage = "none", addEllipses = FALSE, ellipse.level = 0.95,
    col.ind = "black", col.ind.sup = "blue", alpha.ind = 1,
    col.var = "steelblue", alpha.var = 1, col.quanti.sup = "blue",
    col.circle = "grey70", 
    select.var = list(name = NULL, cos2 = NULL, contrib= NULL), 
    select.ind = list(name = NULL, cos2 = NULL, contrib = NULL),
    jitter = list(what = "label", width = NULL, height = NULL), ...)
# An alias of fviz_pca_biplot()
fviz_pca(X, ...)

Arguments

Argument Description
X an object of class PCA [FactoMineR]; prcomp and princomp [stats]; dudi and pca [ade4].
axes a numeric vector of length 2 specifying the dimensions to be plotted.
geom a text specifying the geometry to be used for the graph. Allowed values are the combination of c(“point”, “arrow”, “text”). Use “point” (to show only points); “text” to show only labels; c(“point”, “text”) or c(“arrow”, “text”) to show both types.
label a text specifying the elements to be labelled. Default value is “all”. Allowed values are “none” or the combination of c(“ind”, “ind.sup”, “quali”, “var”, “quanti.sup”). “ind” can be used to label only active individuals. “ind.sup” is for supplementary individuals. “quali” is for supplementary qualitative variables. “var” is for active variables. “quanti.sup” is for quantitative supplementary variables.
invisible a text specifying the elements to be hidden on the plot. Default value is “none”. Allowed values are the combination of c(“ind”, “ind.sup”, “quali”, “var”, “quanti.sup”).
labelsize font size for the labels.
pointsize the size of points.
habillage an optional factor variable for coloring the observations by groups. Default value is “none”. If X is a PCA object from FactoMineR package, habillage can also specify the supplementary qualitative variable (by its index or name) to be used for coloring individuals by groups (see ?PCA in FactoMineR).
addEllipses logical value. If TRUE, draws ellipses around the individuals when habillage != “none”.
ellipse.level the size of the concentration ellipse in normal probability.
col.ind,col.var colors for individuals and variables, respectively. Possible values include also : “cos2”, “contrib”, “coord”, “x” or “y”. In this case, the colors for individuals/variables are automatically controlled by their qualities of representation (“cos2”), contributions (“contrib”), coordinates (x^2 + y^2, “coord”), x values (“x”) or y values (“y”). To use automatic coloring (by cos2, contrib, ….), make sure that habillage =“none”.
col.ind.sup color for supplementary individuals.
alpha.ind,alpha.var controls the transparency of individual and variable colors, respectively. The value can variate from 0 (total transparency) to 1 (no transparency). Default value is 1. Possible values include also : “cos2”, “contrib”, “coord”, “x” or “y”. In this case, the transparency for the individual/variable colors are automatically controlled by their qualities (“cos2”), contributions (“contrib”), coordinates (x^2 + y^2 , “coord”), x values(“x”) or y values(“y”). To use this, make sure that habillage =“none”.
select.ind,select.var

a selection of individuals/variables to be drawn. Allowed values are NULL or a list containing the arguments name, cos2 or contrib:

  • name: is a character vector containing individuals/variables to be drawn
  • cos2: if cos2 is in [0, 1], ex: 0.6, then individuals/variables with a cos2 > 0.6 are drawn. if cos2 > 1, ex: 5, then the top 5 individuals/variables with the highest cos2 are drawn.
  • contrib: if contrib > 1, ex: 5, then the top 5 individuals/variables with the highest cos2 are drawn
jitter a parameter used to jitter the points in order to reduce overplotting. It’s a list containing the objects what, width and height (Ex.; jitter = list(what, width, height)). what: the element to be jittered. Possible values are “point” or “p”; “label” or “l”; “both” or “b”. width: degree of jitter in x direction (ex: 0.2). height: degree of jitter in y direction (ex: 0.2).
col.quanti.sup a color for the quantitative supplementary variables.
col.circle a color for the correlation circle.
Arguments to be passed to the function fviz_pca_biplot().

Value

A ggplot2 plot

Examples

Principal component analysis

A principal component analysis (PCA) is performed using the built-in R function prcomp() and iris data:

data(iris)
head(iris)
  Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1          5.1         3.5          1.4         0.2  setosa
2          4.9         3.0          1.4         0.2  setosa
3          4.7         3.2          1.3         0.2  setosa
4          4.6         3.1          1.5         0.2  setosa
5          5.0         3.6          1.4         0.2  setosa
6          5.4         3.9          1.7         0.4  setosa
# The variable Species (index = 5) is removed
# before the PCA analysis
res.pca <- prcomp(iris[, -5],  scale = TRUE)

fviz_pca_ind(): Graph of individuals

# Default plot
fviz_pca_ind(res.pca)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change title and axis labels
fviz_pca_ind(res.pca) +
  labs(title ="PCA", x = "PC1", y = "PC2")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change axis limits by specifying the min and max
fviz_pca_ind(res.pca) +
   xlim(-4, 4) + ylim (-4, 4)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Use text only
fviz_pca_ind(res.pca, geom="text")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Use points only
fviz_pca_ind(res.pca, geom="point")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change the size of points
fviz_pca_ind(res.pca, geom="point", pointsize = 4)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change point color and theme
fviz_pca_ind(res.pca, col.ind = "blue")+
   theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Control automatically the color of individuals
# using the cos2 or the contributions
# cos2 = the quality of the individuals on the factor map
fviz_pca_ind(res.pca, col.ind="cos2")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Gradient color
fviz_pca_ind(res.pca, col.ind="cos2") +
      scale_color_gradient2(low="white", mid="blue",
      high="red", midpoint=0.6)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change the theme and use only points
fviz_pca_ind(res.pca, col.ind="cos2", geom = "point") +
      scale_color_gradient2(low="white", mid="blue",
      high="red", midpoint=0.6)+ theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Color by the contributions
fviz_pca_ind(res.pca, col.ind="contrib") +
      scale_color_gradient2(low="white", mid="blue",
      high="red", midpoint=4)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Control the transparency of the color by the
# contributions
fviz_pca_ind(res.pca, alpha.ind="contrib") +
     theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Color individuals by groups
fviz_pca_ind(res.pca, label="none", habillage=iris$Species)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Add ellipses
p <- fviz_pca_ind(res.pca, label="none", habillage=iris$Species,
             addEllipses=TRUE, ellipse.level=0.95)
print(p)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change group colors using RColorBrewer color palettes
p + scale_color_brewer(palette="Dark2") +
     theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

p + scale_color_brewer(palette="Paired") +
     theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

p + scale_color_brewer(palette="Set1") +
     theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change color manually
p + scale_color_manual(values=c("#999999", "#E69F00", "#56B4E9"))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select and visualize individuals with cos2 > 0.96
fviz_pca_ind(res.pca, select.ind = list(cos2 = 0.96))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select the top 20 according the cos2
fviz_pca_ind(res.pca, select.ind = list(cos2 = 20))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select the top 20 contributing individuals
fviz_pca_ind(res.pca, select.ind = list(contrib = 20))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select by names
fviz_pca_ind(res.pca,
select.ind = list(name = c("23", "42", "119")))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

fviz_pca_var(): Graph of variables

# Default plot
fviz_pca_var(res.pca)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Use points and text
fviz_pca_var(res.pca, geom = c("point", "text"))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change color and theme
fviz_pca_var(res.pca, col.var="steelblue")+
 theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Control variable colors using their contributions
fviz_pca_var(res.pca, col.var="contrib")+
 scale_color_gradient2(low="white", mid="blue",
           high="red", midpoint=96) +
 theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Control the transparency of variables using their contributions
fviz_pca_var(res.pca, alpha.var="contrib") +
   theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select and visualize variables with cos2 >= 0.96
fviz_pca_var(res.pca, select.var = list(cos2 = 0.96))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select the top 3 contributing variables
fviz_pca_var(res.pca, select.var = list(contrib = 3))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select by names
fviz_pca_var(res.pca,
   select.var= list(name = c("Sepal.Width", "Petal.Length")))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

fviz_pca_biplot(): Biplot of individuals of variables

fviz_pca_biplot(res.pca)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Keep only the labels for variables
fviz_pca_biplot(res.pca, label ="var")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Keep only labels for individuals
fviz_pca_biplot(res.pca, label ="ind")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Hide variables
fviz_pca_biplot(res.pca, invisible ="var")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Hide individuals
fviz_pca_biplot(res.pca, invisible ="ind")

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Control automatically the color of individuals using the cos2
fviz_pca_biplot(res.pca, label ="var", col.ind="cos2") +
       theme_minimal()

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Change the color by groups, add ellipses
fviz_pca_biplot(res.pca, label="var", habillage=iris$Species,
               addEllipses=TRUE, ellipse.level=0.95)

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

# Select the top 30 contributing individuals
fviz_pca_biplot(res.pca, label="var",
               select.ind = list(contrib = 30))

fviz_pca: Quick Principal Component Analysis data visualization - R software and data mining

Infos

This analysis has been performed using R software (ver. 3.2.1) and factoextra (ver. 1.0.3)


Enjoyed this article? I’d be very grateful if you’d help it spread by emailing it to a friend, or sharing it on Twitter, Facebook or Linked In.

Show me some love with the like buttons below... Thank you and please don't forget to share and comment below!!
Avez vous aimé cet article? Je vous serais très reconnaissant si vous aidiez à sa diffusion en l'envoyant par courriel à un ami ou en le partageant sur Twitter, Facebook ou Linked In.

Montrez-moi un peu d'amour avec les like ci-dessous ... Merci et n'oubliez pas, s'il vous plaît, de partager et de commenter ci-dessous!





This page has been seen 268071 times