R_ggplot2_Package

10/16/2024

Data Visualization in R

1 - Base R Graphic Tools

Most people use the base R built-in graphic tools as the starting point for data visualization, which base
R already provides a great set of graphic tools. Here we are going to demonstrate the use of these graphic
tools and later compare it to the more sophisticated ggplot2 package. In this demonstration, we are using
the ‘diamonds’ data set from ggplot2 package.
Histogram is the most fundamental graphic tools to show the frequency distribution of the numerical data.
Base R provides this very easy way to plot the histogram with hist() function. Suppose we are plotting the
histogram for the numerical column “carat” from the diamonds tbl.

library(ggplot2)
data(diamonds)
head(diamonds)

## # A tibble: 6 x 10
## carat cut color clarity depth table price x y z
## <dbl> <ord> <ord> <ord> <dbl> <dbl> <int> <dbl> <dbl> <dbl>
## 1 0.23 Ideal E SI2 61.5 55 326 3.95 3.98 2.43
## 2 0.21 Premium E SI1 59.8 61 326 3.89 3.84 2.31
## 3 0.23 Good E VS1 56.9 65 327 4.05 4.07 2.31
## 4 0.29 Premium I VS2 62.4 58 334 4.2 4.23 2.63
## 5 0.31 Good J SI2 63.3 58 335 4.34 4.35 2.75
## 6 0.24 Very Good J VVS2 62.8 57 336 3.94 3.96 2.48

# Create histogram with base R

hist(diamonds$carat, main='Carat Histogram', xlab='Carat')

1
 Carat Histogram
15000
Frequency

10000
5000
0

0 1 2 3 4 5

Carat

Scatter plot is a very useful tool to compare two sets of numerical data. We can use “formula” to plot the
diamond price against carat.

plot(price ~ carat, data=diamonds)

2
 15000
10000
price

5000
0

1 2 3 4 5

carat

Note: In the formula, price is the y value and carat is the x value.
It is not necessary to use “formula” for a scatter plot,

plot(diamonds$carat, diamonds$price)

3
 15000
diamonds$price

10000
5000
0

1 2 3 4 5

diamonds$carat

Note: the first argument in the plot() function is the x value and the second argument is the y value.
Box plot is probably the first graphical tool learned in fundamental statistic class. It is used to depict the
range, mean, median, and quartiles of the numerical data. We can use the boxplot() function to depict the
numerical data ‘carat’

boxplot(diamonds$carat)

4
 5
4
3
2
1

2 - ggplot2 Package

Base R already provides a powerful graphic tools for the average R users. However, when we want to modify
some specific features in the graph, the base R graphic tools requires more lines of code. ggplot2 package
offers a structure of coding for achieving the complex modification of graphic features. In some cases, adding
some additional features to a graph may take 30 additional lines of code to run in base R, but may only
require 1 line of code using ggplot2 package.
We start with taking the data as the first argument in ggplot() function and build on top of it with “+”.
Think of “+” is like adding an additional layer to the plot for different feature added. The most important
thing to add into the new layer is aes(), represents “aesthetic” adjustment.

Histogram with ggplot2

Using ggplot2 for plotting the histogram for numerical data ‘Carat’ in the diamonds data set.

ggplot(data=diamonds) + geom_histogram(aes(x=carat))

## ‘stat_bin()‘ using ‘bins = 30‘. Pick better value with ‘binwidth‘.

5
 15000

10000
count

5000

0 1 2 3 4 5
carat

Similar to histogram, we can easily plot the density distribution with geom_density().

ggplot(data=diamonds) + geom_density(aes(x=carat), fill='grey50')

6
 1.5

1.0
density

0.5

0.0

0 1 2 3 4 5
carat

Note: Histogram divide the continuous variable into groups (x-axis) and gives the frequency (y-axis) in each
group. Density distribution is evaluated at any given point on the continuous variable (x-axis)

Scatter Plot with ggplot2

In this section, we are not only demonstrating how to use scatter plot with ggplot2 package, we also try to
demonstrate the simplicity for adding some useful features to our graph. Plot the diamond price and carat
within a scatter plot. This time we are including the aes() in the ggplot() function, but not in the geom()

# Create scatter plot with ggplot2

ggplot(diamonds, aes(x=carat, y=price)) + geom_point()

7
 15000
price

10000

5000

0
0 1 2 3 4 5
carat

# We are going to repeat using "ggplot(diamonds, aes(x=carat, y=price))", so we save it into a variable
g <- ggplot(diamonds, aes(x=carat, y=price))

# Suppose we would like to identify the colors of the diamond in the graph with different color codes.
g + geom_point(aes(color=color))

8
 15000
color
D
E
F
price

10000
G
H
I
J
5000

0
0 1 2 3 4 5
carat

Note: The argument color= is to define what color to use for the data points in the graph and we define the
color will be different according to the value under column “color” in the data set.
ggplot2 can also create layers of plot with functions like facet_wrap() or facet_grid(). For instance, if we
are interested to split each color grade into its own price vs. carat scatter plot.

g + geom_point(aes(color=color)) + facet_wrap(~color)

9
 D E F

15000

10000

5000

0 color
G H I D
E
15000
F
price

10000
G
5000
H
0
0 1 2 3 4 5 0 1 2 3 4 5 I
J
J
15000

10000

5000

0
0 1 2 3 4 5
carat

# Or, if we want to splite the scatter plots into different combinations of "cut" and "clarity" grade.
g + geom_point(aes(color=color)) + facet_grid(cut~clarity)

10
 I1 SI2 SI1 VS2 VS1 VVS2 VVS1 IF

15000

Fair
10000
5000
0

15000
color

Good
10000
5000 D
0
E

Very Good
15000 F
price

10000
5000 G
0 H

Premium
15000 I
10000
J
5000
0

15000

Ideal
10000
5000
0
012345012345012345012345012345012345012345012345
carat

# Note: Both facet_wrap() and facet_grid() can be applied to any geom() functions.
# e.g. We can use facet_wrap() to plot the histrogram for carat in different color grade.
ggplot(diamonds, aes(x=carat)) + geom_histogram() + facet_wrap(~color)

## ‘stat_bin()‘ using ‘bins = 30‘. Pick better value with ‘binwidth‘.

11
 D E F

3000

2000

1000

0
G H I

3000
count

2000

1000

0
0 1 2 3 4 5 0 1 2 3 4 5
J

3000

2000

1000

0
0 1 2 3 4 5
carat

Box Plot with ggplot2

ggplot2 provides a more complete and intuitive features to the box plot graphic. Plotting the box plot for
carat

# Create box plot with ggplot2

ggplot(diamonds, aes(y=carat, x=1)) + geom_boxplot()

12
 5

3
carat

0
0.6 0.8 1.0 1.2 1.4
x

Note: Unlike the base R boxplot(), ggplot2 require to define the x-axis even for a one dimensional plot.
Therefore, we put in the value 1 for x.
In ggplot2, we can easily extend the boxplot into different classes by defining a categorical data. We can
divide the box plot for carat by the cut grade.

# Create boxplot by different cut grade

ggplot(diamonds, aes(y=carat, x=cut)) + geom_boxplot()

13
 5

3
carat

0
Fair Good Very Good Premium Ideal
cut

# In a more specific case, we can change the box plot into a violin plot
ggplot(diamonds, aes(y=carat, x=cut)) + geom_violin()

14
 5

3
carat

0
Fair Good Very Good Premium Ideal
cut

Line Graph with ggplot2

Line graph is usually applied to express the time series data. We are using the economics data set from
ggplot2 package to demonstrate the use of line graph with date. Plot the population data with the given
date.

# Create line graph with ggplot2

ggplot(economics, aes(x=date, y=pop)) + geom_line()

15
 325000

300000

275000
pop

250000

225000

200000

1970 1980 1990 2000 2010

date

Suppose we are trying to break the data into year and plot into the same graph. We can use the lubridate
package to help us on this task.

library(lubridate)

## Warning: package ’lubridate’ was built under R version 4.3.3

##
## Attaching package: ’lubridate’

## The following objects are masked from ’package:base’:

##
## date, intersect, setdiff, union

# Build year and month variables

economics$year <- year(economics$date)
# The argument label=TRUE means to return month by its name, not a numeric value
economics$month <- month(economics$date, label=TRUE)
# Let's save the data in or beyond 2000 to an object "econ2000"
econ2000 <- economics[which(economics$year >= 2000),]

# We can use scales package to give better format to the graph

library(scales)

# Build the base of the line graph

16
g <- ggplot(econ2000, aes(x=month, y=pop))
# Define different colors for different years.
# Use argument group to group the data into different year.
g <- g +geom_line(aes(color=factor(year), group=year))
# Define the color by the given "Year"
g <- g + scale_color_discrete(name="Year")
# Format the y-axis
g <- g + scale_y_continuous(labels=comma)
# Adding the title, x-axis, and y-axis to the graph
g <- g + labs(title="Population Growth", x="Month", y="Population")
# Plot the graph with all the added features
g

Population Growth Year

320,000 2000
2001
2002
2003
310,000
2004
2005
Population

2006
2007
300,000
2008
2009
2010
290,000 2011
2012
2013
2014
280,000
2015
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Themes in ggplot2

Another big advantage using ggplot2 is that the package provides users variety of theme options, such as
“The Economist”, “Excel”, “Edward Tufte”, and “The Wall Street Journal”. To access the themes, we need
to load the ggthemes library package.

# [Link]("ggthemes")
library(ggthemes)

## Warning: package ’ggthemes’ was built under R version 4.3.3

17
# e.g. Applying differen themes on a scatter plot.
g2 <- ggplot(diamonds, aes(x=carat, y=price)) +
geom_point(aes(color=color))

g2 + theme_economist() + scale_color_economist()

D F H J
color
E G I

15000
price

10000

5000

0 1 2 3 4 5
carat

g2 + theme_excel() + scale_color_excel()

18
 15000
color
D
E
F
price

10000
G
H
I
J
5000

0
0 1 2 3 4 5
carat

g2 + theme_tufte()

19
 15000
color
D
E
F
price

10000
G
H
I
J
5000

0
0 1 2 3 4 5
carat

g2 + theme_wsj()

20
 D F H J
color E G I

15000

10000

5000

0
0 1 2 3 4 5

21