SLIDE 1
Graphics in R
STAT 133 Gaston Sanchez
Department of Statistics, UC–Berkeley gastonsanchez.com github.com/gastonstat/stat133 Course web: gastonsanchez.com/stat133
R Graphics
2
Graphics in R STAT 133 Gaston Sanchez Department of Statistics, - - PowerPoint PPT Presentation
Graphics in R STAT 133 Gaston Sanchez Department of Statistics, - - PowerPoint PPT Presentation
Graphics in R STAT 133 Gaston Sanchez Department of Statistics, UCBerkeley gastonsanchez.com github.com/gastonstat/stat133 Course web: gastonsanchez.com/stat133 R Graphics 2 Understanding Graphics in R 2 main graphics systems
SLIDE 2
SLIDE 3
Understanding Graphics in R 2 main graphics systems "graphics" & "grid"
3
SLIDE 4
Basics of Graphics in R
Graphics Systems
◮ "graphics" and "grid" are the two main graphics
systems in R
◮ "graphics" is the traditional system, also referred to as
base graphics
◮ "grid" prodives low-level functions for programming
plotting functions
4
SLIDE 5
Basics of Graphics in R
Graphics Engine
◮ Underneath "graphics" and "grid" there is the package
"grDevices"
◮ "grDevices" is the graphics engine in R ◮ It provides the graphics devices and support for colors and
fonts
5
SLIDE 6
grid graphics
grDevices
maps diagram plotrix ggplot2 lattice
tikzDevice
JavaGD Cairo
6
SLIDE 7
Basics of Graphics in R
Package "graphics"
The package "graphics" is the traditional system; it provides functions for complete plots, as well as low-level facilities. Many other graphics packages are built on top of graphics like "maps", "diagram", "pixmap", and many more.
7
SLIDE 8
Understanding Graphics in R
Package "grid"
The "grid" package does not provide functions for drawing complete plots. "grid" is not used directly to produce statistical plots. Instead, it is used to build other graphics packages like "lattice" or "ggplot2".
8
SLIDE 9
In this course
◮ In this course we’ll focus on the packages "graphics" and
"ggplot2"
◮ "graphics" is the traditional plotting system in R, and
many functions and packages are built on top of it.
◮ "ggplot2" excels at providing graphics for visualizing
multivariate data sets —in data.frame format—, while taking care of many issues for superior visual displays.
9
SLIDE 10
R Graphics by Paul Murrell
10
SLIDE 11
Some Resources
◮ R Graphics by Paul Murrell
book and webpage
◮ R Graphics Cookbook by Winston Chang
http://www.cookbook-r.com/Graphs/
◮ ggplot2: Elegant Graphics for Data Analysis by
Hadley Wickham
◮ R Graphs Cookbook by Hrishi Mittal ◮ Graphics for Statistics and Data Analysis with R by
Kevin Keen
11
SLIDE 12
Traditional (Base) Graphics
12
SLIDE 13
Base Graphics in R
Types of graphics functions
Graphics functions can be divided into two main types:
◮ high-level functions produce complete plots, e.g.
barplot(), boxplot(), dotchart()
◮ low-level functions add further output to an existing plot,
e.g. text(), points(), legend()
13
SLIDE 14
The plot() function
◮ plot() is the most important high-level function in
traditional graphics
◮ The first argument to plot() provides the data to plot ◮ The provided data can take different forms: e.g. vectors,
factors, matrices, data frames.
◮ To be more precise, plot() is a generic function ◮ You can create your own plot() method function 14
SLIDE 15
Basic Plots with plot()
In its basic form, we can use plot() to make graphics of:
◮ one single variable ◮ two variables ◮ multiple variables 15
SLIDE 16
Plots of One Variable
16
SLIDE 17
High-level graphics of a single variable
Function Data Description plot() numeric scatterplot plot() factor barplot plot() 1-D table barplot numeric can be either a vector or a 1-D array (e.g. row or column from a matrix)
17
SLIDE 18
One variable objects
Vector / Factor row (data.frame) row (matrix) 1-D table column (data.frame) column (matrix)
18
SLIDE 19
plot() of one variable
# plot numeric vector num_vec <- (c(1:10))^2 plot(num_vec) # plot factor set.seed(4) abc <- factor(sample(c('A', 'B', 'C'), 20, replace = TRUE)) plot(abc) # plot 1D-table abc_table <- table(abc) plot(abc_table)
19
SLIDE 20
plot() of one variable
- 2
4 6 8 10 20 40 60 80 100 Index num_vec A B C 2 4 6 8 2 4 6 8 abc abc_table A B C
20
SLIDE 21
More high-level graphics of a single variable
Function Data Description barplot() numeric barplot pie() numeric pie chart dotchart() numeric dotplot boxplot() numeric boxplot hist() numeric histogram stripchart() numeric 1-D scatterplot stem() numeric stem-and-leaf plot
21
SLIDE 22
Plots of one variable
# barplot numeric vector barplot(num_vec) # pie chart pie(1:3) # dot plot dotchart(num_vec)
22
SLIDE 23
Plots of one variable
20 40 60 80 100 1 2 3
- 20
40 60 80 100
23
SLIDE 24
Plots of one variable
# barplot numeric vector boxplot(num_vec) # pie chart hist(num_vec) # dot plot stripchart(num_vec) # stem-and-leaf stem(num_vec)
24
SLIDE 25
boxplot()
# boxplot boxplot(iris$Sepal.Length)
4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
25
SLIDE 26
hist()
# histogram hist(iris$Sepal.Length)
Histogram of iris$Sepal.Length
iris$Sepal.Length Frequency 4 5 6 7 8 5 10 15 20 25 30
26
SLIDE 27
Test your knowledge
What option does not apply to histograms:
A) adjacent bars (no gaps) B) area of bars indicate proportions C) bins of equal length D) bars can be reordered
27
SLIDE 28
stripchart()
# strip-chart (1-D scatter plot) # (for small sample sizes) stripchart(num_vec)
20 40 60 80 100
28
SLIDE 29
stem()
# stem-and-leaf plot # (for small sample sizes) stem(num_vec) ## ## The decimal point is 1 digit(s) to the right of the | ## ## 0 | 1496 ## 2 | 56 ## 4 | 9 ## 6 | 4 ## 8 | 1 ## 10 | 0
29
SLIDE 30
Kernel Density Curve
◮ Surprisingly, R does not have a specific function to plot
density curves
◮ R does have the density() function which computes a
kernel density estimate
◮ We can pass a "density" object to plot() in order to
get a density curve.
30
SLIDE 31
Kernel Density Curve
# kernel density curve dens <- density(num_vec) plot(dens)
−50 50 100 150 0.000 0.004 0.008
density.default(x = num_vec)
N = 10 Bandwidth = 19.41 Density
31
SLIDE 32
Test your knowledge
What type of plot is based on the five-number summary
A) bar chart B) box plot C) histogram D) scatterplot
32
SLIDE 33
Plots of Two Variables
33
SLIDE 34
High-level graphics of two variables
Function Data Description plot() numeric, numeric scatterplot plot() numeric, factor stripcharts plot() factor, numeric boxplots plot() factor, factor spineplot plot() 2-column numeric matrix scatterplot plot() 2-column numeric data.frame scatterplot plot() 2-D table mosaicplot
34
SLIDE 35
Two variable objects
2-D table (frequency or crosstable) 2-column (numeric data.frame) 2-column (numeric matrix) 2 numeric vectors num vector, factor factor, num vector 2 factors
35
SLIDE 36
Plots of two variables
# plot numeric, numeric plot(iris$Petal.Length, iris$Sepal.Length) # plot numeric, factor plot(iris$Petal.Length, iris$Species) # plot factor, numeric plot(iris$Species, iris$Petal.Length) # plot factor, factor plot(iris$Species, iris$Species)
36
SLIDE 37
Plots of two variables
# plot numeric, numeric plot(iris$Petal.Length, iris$Sepal.Length)
- ●
- ●
- ●
- ●
- ●
- 1
2 3 4 5 6 7 4.5 5.5 6.5 7.5 iris$Petal.Length iris$Sepal.Length 37
SLIDE 38
Plots of two variables
# plot numeric, factor plot(iris$Petal.Length, iris$Species)
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ● ●
- ● ●
- ●
- ●
- ●
- ● ●
- 1
2 3 4 5 6 7 1.0 1.5 2.0 2.5 3.0 iris$Petal.Length iris$Species 38
SLIDE 39
Plots of two variables
# plot factor, numeric plot(iris$Species, iris$Petal.Length)
- setosa
versicolor virginica 1 2 3 4 5 6 7 39
SLIDE 40
Plots of two variables
# plot factor, factor plot(iris$Species, iris$Species)
x y setosa versicolor virginica setosa versicolor virginica 0.0 0.2 0.4 0.6 0.8 1.0
40
SLIDE 41
Plots of two variables
# some fake data set.seed(1) # hair color hair <- factor( sample(c('blond', 'black', 'brown'), 100, replace = TRUE)) # eye color eye <- factor( sample(c('blue', 'brown', 'green'), 100, replace = TRUE))
41
SLIDE 42
Plots of two variables
# plot factor, factor plot(hair, eye)
x y black blond brown blue brown green 0.0 0.2 0.4 0.6 0.8 1.0
42
SLIDE 43
More high-level graphics of two variables
Function Data Description sunflowerplot() numeric, numeric sunflower scatterplot smoothScatter() numeric, numeric smooth scatterplot boxplot() list of numeric boxplots barplot() matrix stacked / side-by-side barplot dotchart() matrix dotplot stripchart() list of numeric stripcharts spineplot() numeric, factor spinogram cdplot() numeric, factor conditional density plot fourfoldplot() 2x2 table fourfold display assocplot() 2-D table association plot mosaicplot() 2-D table mosaic plot
43
SLIDE 44
Plots of two variables
# sunflower plot (numeric, numeric) sunflowerplot(iris$Petal.Length, iris$Sepal.Length)
1 2 3 4 5 6 7 4.5 5.5 6.5 7.5 iris$Petal.Length iris$Sepal.Length
- ●
- 44
SLIDE 45
Plots of two variables
# smooth scatter plot (numeric, numeric) smoothScatter(iris$Petal.Length, iris$Sepal.Length)
1 2 3 4 5 6 7 4.5 5.5 6.5 7.5 iris$Petal.Length iris$Sepal.Length 45
SLIDE 46
Plots of two variables
# boxplots (numeric, numeric) boxplot(iris$Petal.Length, iris$Sepal.Length)
1 2 1 2 3 4 5 6 7 8 46
SLIDE 47
Plots of two variables
m <- matrix(1:8, 4, 2) # barplot (numeric matrix) barplot(m)
5 10 15 20 25 47
SLIDE 48
Plots of two variables
m <- matrix(1:8, 4, 2) # barplot (numeric matrix) barplot(m, beside = TRUE)
2 4 6 8 48
SLIDE 49
Plots of two variables
# conditional density plot (numeric, factor) cdplot(iris$Petal.Length, iris$Species)
iris$Petal.Length iris$Species 2 3 4 5 6 setosa virginica 0.0 0.2 0.4 0.6 0.8 1.0 49
SLIDE 50
Two categorical variables: frequency table
# 2-D table (HairEyeColor data) x <- margin.table(HairEyeColor, c(1, 2)) x ## Eye ## Hair Brown Blue Hazel Green ## Black 68 20 15 5 ## Brown 119 84 54 29 ## Red 26 17 14 14 ## Blond 7 94 10 16
50
SLIDE 51
Plots of two categorical variables
# mosaic plot (2-D table) mosaicplot(x, main = "Relation between hair and eye color")
Relation between hair and eye color
Hair Eye
Black Brown Red Blond Brown Blue Hazel Green
51
SLIDE 52
Plots of two categorical variables
# association plot (2-D table) assocplot(x, main = "Relation between hair and eye color")
Black Brown Red Blond Green Blue Brown
Relation between hair and eye color
Hair Eye 52
SLIDE 53
Plots of Multiple Variables
53
SLIDE 54
High-level graphics of multiple variables
Function Data Description plot() data frame scatterplot matrix pairs() matrix scatterplot matrix matplot() matrix scatterplot stars() matrix star plots image() numeric, numeric, numeric image plot contour() numeric, numeric, numeric contour plot filled.contour() numeric, numeric, numeric filled contour plot persp() numeric, numeric, numeric 3-D surface symbols() numeric, numeric, numeric symbols scatterplot mosaicplot() N-D table mosaic plot
54
SLIDE 55
Plots of multiple variables
# scatter plot matrix (data frame) plot(iris[ , 1:4]) ## Warning: closing unused connection 5 (http://gastonsanchez.com/education.csv)
Sepal.Length
2.0 3.0 4.0
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 0.5
1.5 2.5 4.5 6.0 7.5
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 2.0
3.0 4.0
- ●
- ●
- ●
- Sepal.Width
- ●
- ●
- ● ●
- ●
- ● ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ● ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- Petal.Length
1 3 5 7
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 0.5
1.5 2.5
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- Petal.Width
55
SLIDE 56
Plots of multiple variables
# scatter plot matrix (data frame) pairs(iris[ , 1:4])
Sepal.Length
2.0 3.0 4.0
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 0.5
1.5 2.5 4.5 6.0 7.5
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 2.0
3.0 4.0
- ●
- ●
- ●
- Sepal.Width
- ●
- ●
- ● ●
- ●
- ● ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ● ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- Petal.Length
1 3 5 7
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 4.5
6.0 7.5 0.5 1.5 2.5
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- ●
- 1
3 5 7
- ●
- ●
- ●
- Petal.Width
56
SLIDE 57
Plots of multiple variables
# scatter plot matrix (data frame) matplot(iris[ , 1:4])
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 50 100 150 2 4 6 8 iris[, 1:4] 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
57
SLIDE 58
Plots of multiple variables
# star plot (data frame) stars(iris[ , 1:4])
58
SLIDE 59
Plots of multiple variables
# color image (matrix) image(t(volcano)[ncol(volcano):1, ])
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
59
SLIDE 60
Plots of multiple variables
# display of Maunga Whau volcano x <- 10*(1:nrow(volcano)) y <- 10*(1:ncol(volcano)) image(x, y, volcano, col = terrain.colors(100), axes = FALSE) contour(x, y, volcano, levels = seq(90, 200, by = 5), add = TRUE, col = "peru") axis(1, at = seq(100, 800, by = 100)) axis(2, at = seq(100, 600, by = 100)) box() title(main = "Maunga Whau Volcano", font.main = 4)
60
SLIDE 61
Plots of multiple variables
x y
95 1 100 1 105 1 5 105 110 110 110 1 1 115 115 115 120 125 130 135 140 145 150 1 5 5 155 1 6 160 165 165 1 7 170 175 180 1 8 5 190
100 200 300 400 500 600 700 800 100 200 300 400 500 600
Maunga Whau Volcano
61
SLIDE 62
Plots of multiple variables
# mosaic plot of N-D tables mosaicplot(HairEyeColor) HairEyeColor
Hair Eye
Black Brown Red Blond Brown Blue Hazel Green MaleFemale Male Female Male Female Male Female
62
SLIDE 63
Plots of multiple variables
# symbols scatter plots symbols(iris[, 1], iris[, 2], circles = iris[, 3]/100, inches = FALSE) 4 5 6 7 8 2.0 3.0 4.0 iris[, 1] iris[, 2]
- ●
- ●
- ●
- 63
SLIDE 64
Graphics Parameters
Graphics Functions and Arguments
◮ Plot functions usually come with various arguments ◮ Typically, the first argument(s) is the data object(s) to be
plotted
◮ Most of the other arguments have default options ◮ Graphic arguments have a consisting naming convention,
but there will always be some exception
64
SLIDE 65
Graphical Parameters
Graphical Arguments
◮ Some arguments are specific to a function (e.g. horiz or
beside in barplot())
◮ Other arguments are more general (e.g. col, xlab, ylab) ◮ General graphical parameters are listed in the
documentation of the function par()
◮ See ?par for more information 65
SLIDE 66
Graphics in R
How to choose a graphics approach?
◮ look first for an existing function that does what you want
—or something similar to what you want (don’t reivent the wheel!)
◮ Existing plotting functions can be combined and
customized by using optional arguments or graphical parameters
◮ For exploratory data analysis (quick and dirty) the plotting
functions in "graphics" is a good option
66