introduction
prerequisites
data
ggscatmat()
ggpairs()
non-ggplot packages
exercises
references
A scatterplot matrix (or pairs plot) is a graph design for visualizing three or more quantitative variables and (possibly) categorical variables. The scatterplot matrix is a grid of scatterplots showing the bivariate relationships between all pairs of variables (Emerson and others, 2013).
Data characteristics
- 3 or more quantitative variables
- 1 or more categorical variables (optional). If you have a categorical variable, some packages cannot be used.
- A key variable if data are not coordinatized
Graph characteristics
- A matrix of scatterplots
- The variable names are the labels of the rows and the columns of the matrix
- Optional: loess or other smooth fit
- Optional: the matrix diagonal shows a statistical summary of the variable
- Optional: pair-wise correlation coefficients
D6 Multivariate data and graph requirements
Project setup
- Start every work session by launching the RStudio Project file for
the course, e.g.,
portfolio.Rproj - Ensure your project directory structure satisfies the course requirements
Ensure you have installed the following packages. See install packages for instructions if needed.
- tidyverse: The ‘tidyverse’ is a set of packages that work in harmony because they share common data representations and ‘API’ design. This package is designed to make it easy to install and load multiple ‘tidyverse’ packages in a single step. Learn more about the ‘tidyverse’ at https://tidyverse.org.
- graphclassmate: An R package with companion materials for a course in data visualization. The package provides data sets structured for a variety of graph types plus a ggplot2 theme.
- GGally: The R package ‘ggplot2’ is a plotting system based on the grammar of graphics. ‘GGally’ extends ‘ggplot2’ by adding several functions to reduce the complexity of combining geometric objects with transformed data. Some of these functions include a pairwise plot matrix, a two group pairwise plot matrix, a parallel coordinates plot, a survival plot, and several functions to plot networks.
- Sleuth2: Data sets from Ramsey, F.L. and Schafer, D.W. (2002), “The Statistical Sleuth: A Course in Methods of Data Analysis (2nd ed)”, Duxbury.
- car: Functions to Accompany J. Fox and S. Weisberg, An R Companion to Applied Regression, Third Edition, Sage, in press.
- gclus: Orders panels in scatterplot matrices and parallel coordinate displays by some merit index. Package contains various indices of merit, ordering functions, and enhanced versions of pairs and parcoord which color panels according to their merit level.
- gpairs: Produces a generalized pairs (gpairs) plot.
Scripts to initialize
explore/ 0801-scatterplot-matrix-explore.R
And start the file with a minimal header
# your name
# date
# load packages
library("tidyverse")
library("graphclassmate")Duplicate the lines of code in the session one chunk at a time. Save, Source, and compare your results to the results shown.
Open the explore script you initialized earlier. Load the package that
has the data. These data are measurements made of genuine and
counterfeit Swiss bank notes. To learn more about the data set, open its
help page by running ? bank. All dimensions are in mm.
library("gclus")
data(bank, package = "gclus")
glimpse(bank)
#> Observations: 200
#> Variables: 7
#> $ Status <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...
#> $ Length <dbl> 214.8, 214.6, 214.8, 214.8, 215.0, 215.7, 215.5, 214....
#> $ Left <dbl> 131.0, 129.7, 129.7, 129.7, 129.6, 130.8, 129.5, 129....
#> $ Right <dbl> 131.1, 129.7, 129.7, 129.6, 129.7, 130.5, 129.7, 129....
#> $ Bottom <dbl> 9.0, 8.1, 8.7, 7.5, 10.4, 9.0, 7.9, 7.2, 8.2, 9.2, 7....
#> $ Top <dbl> 9.7, 9.5, 9.6, 10.4, 7.7, 10.1, 9.6, 10.7, 11.0, 10.0...
#> $ Diagonal <dbl> 141.0, 141.7, 142.2, 142.0, 141.8, 141.4, 141.6, 141....The status variable is an integer, where 0 = a genuine bank note and 1 = a counterfeit bank note. To condition the graphs by status, we convert the variable to a factor with the levels “genuine” and “counterfeit.”
bank <- bank %>%
mutate(Status = factor(Status, labels = c("genuine", "counterfeit"))) %>%
glimpse()
#> Observations: 200
#> Variables: 7
#> $ Status <fct> genuine, genuine, genuine, genuine, genuine, genuine,...
#> $ Length <dbl> 214.8, 214.6, 214.8, 214.8, 215.0, 215.7, 215.5, 214....
#> $ Left <dbl> 131.0, 129.7, 129.7, 129.7, 129.6, 130.8, 129.5, 129....
#> $ Right <dbl> 131.1, 129.7, 129.7, 129.6, 129.7, 130.5, 129.7, 129....
#> $ Bottom <dbl> 9.0, 8.1, 8.7, 7.5, 10.4, 9.0, 7.9, 7.2, 8.2, 9.2, 7....
#> $ Top <dbl> 9.7, 9.5, 9.6, 10.4, 7.7, 10.1, 9.6, 10.7, 11.0, 10.0...
#> $ Diagonal <dbl> 141.0, 141.7, 142.2, 142.0, 141.8, 141.4, 141.6, 141....In the following graphs, I use the scatterplot matrix function from several packages. Each has a somewhat different look and feel. Some are easier than others to edit.
I’ll be using the same color scheme in each case, so I’ll assign a couple of color vectors here.
my_color <- c(rcb("dark_BG"), rcb("dark_Br"))
my_fill <- c(rcb("light_BG"), rcb("light_Br"))
my_title <- "Comparing Swiss banknote dimensions (mm)"Package GGally extends ggplot2.
ggscatmat() treats continuous variables only, though a categorical
variable can be mapped to the color aesthetic.
library("GGally")
ggscatmat(bank, columns = 2:7)
Include the Status category and edit the aesthetics. GGally is an
extension of ggplot2, so its functions are generally compatible with
ggplot2 functions such as scale_color_manual(), labs(), and
theme().
ggscatmat(bank, columns = 2:7, color = "Status") +
geom_point(size = 1, alpha = 0.1, na.rm = TRUE) +
scale_x_continuous(breaks = seq(0, 300, 1)) +
scale_y_continuous(breaks = seq(0, 300, 1)) +
scale_color_manual(values = my_color) +
labs(title = my_title) +
theme(legend.position = "right",
panel.spacing = unit(1, "mm"),
axis.text.x = element_text(angle = 90, hjust = 1))
Package GGally extends ggplot2.
ggpairs() is the most general of the scatterplot matrix functions,
permitting a lot of detailed control—and is thus more complex. It treats
both quantitative and categorical variables in the panels. .
library("GGally")
ggpairs(bank, columns = 2:7)
Include the Status category and edit the aesthetics.
pm <- ggpairs(bank, columns = 2:7,
mapping = ggplot2::aes(color = Status, fill = Status),
title = my_title,
legend = 1,
upper = list(continuous = wrap("cor", size = 2.5))) +
theme(legend.position = "right",
panel.spacing = unit(1, "mm"),
axis.text.x = element_text(angle = 90, hjust = 1))
# loop through each panel to edit colors
for(i in 1:pm$nrow) {
for(j in 1:pm$ncol){
pm[i, j] <- pm[i, j] +
scale_fill_manual(values = my_fill) +
scale_color_manual(values = my_color)
}}
# index to the panels I want to edit alpha
row_col_index <- wrapr::build_frame(
"row", "col" |
1, 1 |
2, 2 |
3, 3 |
4, 4 |
5, 5 |
6, 6
)
# add alpha to the density plots on the diagonal
for(i in 1:nrow(row_col_index)) {
ii <- row_col_index$row[i]
jj <- row_col_index$col[i]
p <- pm[ii, jj]
p <- p + geom_density(alpha = 0.6)
pm[ii, jj] <- p
}
pm
If the GGally functions suit your needs, you may skip this section.
If not, here are some other packages that create scatterplot matrices.
base R pairs()
Here I use the pairs() function and edit the aesthetics and group by
status using base R syntax.
pairs(~ Length + Left + Right + Bottom + Top + Diagonal,
data = bank,
pch = c(21, 21)[bank$Status],
col = my_color[bank$Status],
bg = my_fill[bank$Status],
gap = 0,
upper.panel = NULL,
cex.labels = 1,
las = 2,
main = my_title
)
par(xpd = NA) # clip to device
legend("topright",
title = "Swiss banknotes",
legend = levels(bank$Status),
col = my_color,
pt.bg = my_fill,
pch = 21,
inset = c(0.2, 0.2),
bty = "n", # no border on legend
cex = 0.8,
y.intersp = 0.75,
title.adj = 0.5) 
par(xpd = FALSE) # return to default
car scatterplotMatrix()
Package car (Companion to Applied Regression), spm() builds on the
base R pairs() function. Data must be numeric.
library("car")
scatterplotMatrix(~ Length + Left + Right + Bottom + Top + Diagonal | Status,
data = bank,
pch = c(16, 3),
cex = 0.75 * c(1, 1),
col = my_color,
cex.labels = 1,
cex.axis = 1,
cex.main = 1,
main = my_title,
use = "pairwise.complete.obs"
)
gpairs gpairs()
Package gpairs. Any combination of quantitative and categorical variables is acceptable.
library("gpairs")
gpairs(bank[ , 2:7],
lower.pars = list(scatter = "points"),
upper.pars = list(scatter = 'stats'),
scatter.pars = list(pch = 16,
size = unit(5, "pt"),
col = my_color[bank$Status],
frame.fill = NULL,
border.col = "gray50"),
stat.pars = list(verbose = FALSE),
gap = 0
)
1. case1202
Script: explore/0801-scatterplot-matrix-case1202-explore.R
Data: case1202 from package Sleuth2
-
Explore: Identify the number of observations and the number and type and class of variables.
-
Carpentry: Select the variables Sex, Senior, Age, Bsal, Sal77. Convert dollars to 1000s and months to years.
-
Design: Create a scatterplot matrix using any of the packages/functions illustrated above. Plot the quantitative variables in the panels and condition by the categorical variable. Attempt to create a loess curve in each panel.
-
What stories do you see in these data?
Answer
Emerson JW, Green WA, Schloerke B, Crowley J, Cook D, Hofmann H and Wickham H (2013) The generalized pair plot. Journal of Computational and Graphical Statistics 22(1), 79–91 doi:10.1080/10618600.2012.694762
Wickham H and Grolemund G (2017) R for Data Science. O’Reilly Media, Inc., Sebastopol, CA https://r4ds.had.co.nz/