Introduction to R

Data visualization in ggplot2

Yu Cheng Hsu

Learning objectives

By the end of this session, you will

• Understand the basic grammar and layered structure of ggplot2
• Identify common graph components and choose an appropriate chart type
• Create simple plots using data, aesthetics, and geometric layers
• Interpret the main elements of a ggplot2 figure

We DO NOT expect:

• Memorize functions in the ggplot2

Background image: Mondrian (1935)

Preamble

About Ray’s part

I am responsible for introducing commonly used libraries in R. So most of the time we will keep using functions that are predefined by others.

Questions that Ray could barely give you an explanation better than AI/documentation:

• “What does this function do” -> Go to find the documentation ?FUNCTION_NAME
• “Why is this named XXXX()” -> The developer named it that way

Competency you need to develop:

• “Why we need to do this?”
• Describe the task so AI can help you generate code
• Comprehend and relate to the code and make your own judgment

Slide conventions

• Use code formatting for special terms and examples, such as This or This are code.

• Code chunk, and it is executable when you open the file in RStudio

# Title: Example code chunk
# This chunk shows the basic structure of an executable R block.

• .qmd version file won’t have the image display but the content is the same
• In-class sketch might or might not be uploaded

Installing packages

Package

• A package is a bundle of functions, data, and documentation. You can import them through

library(package_name)

• while using functions under the package you can use

package_name::function_name()

or

library(package_name)
function_name()

• You can check what packages have been installed, imported to your environment using the tab Packages

Checking packages

Installing package

Golden rule: refer to the documentation of the specific package

• CRAN (the Comprehensive R Archive Network) (Major package channel)

install.packages("ggplot2")
install.packages("tidyverse")

• Bioconductor (Biomedical packages)

BiocManager::install("DESeq2")

• GitHub (code at the developmental stage)

devtools::install_github("tidyverse/ggplot2")

ggplot2

Introduction to ggplot2

About ggplot2

• Visualization library in R for creating elegant and complex visualizations.
• ggplot2 is part of the tidyverse collection of R packages.
• Developed by Hadley Wickham (Wickham 2010). He received the COPSS Presidents’ Award for his contributions to the tidyverse collection.
• Flexible and intuitive once you understand its core concepts.

library(ggplot2)

About the data

iris is a dataset introduced by Ronald Fisher in 1936 that collects iris flower morphology measurements from 3 species

• Setosa,
• Versicolor
• Virginica

it measures the length and width of iris’s sepal and petal

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

Iris flower

ggplot2 example

• What are the elements in a graph ?

Components in the graph

Common components of a figure, figure from Wickham (2016)

Choosing the Visualization

Thinking process and decision flowchart will help you determine which type of graph to use (at least within the scope of this course):

1. What is the purpose of displaying the graph?
2. What types of data are you going to present?

flowchart LR
  A{Data type} -->|Continuous| B{Purpose}
  A{Data type}  -->|Discrete| C{Purpose}

  B{Purpose}-->|Exploration| D((Histogram/Boxplot))
  B{Purpose} -->|Association| E((Scatter plot))
  B{Purpose} -->|Association+time| T((Line plot))

  C{Purpose}-->|Exploration| F((Bar chart))
  C{Purpose} -->|Association| G((Tree map))

Overview of ggplot2 logics

ggplot2 is designed as if you are plotting by hand, so the process looks like

1. Defining coordinate systems
2. Plotting dots, lines, bars, and etc.
3. Adding legend and axis information

Layer features of ggplot2

Layer 1: Coordinate system

First, we define the coordinate system for our iris data in R

The function is ggplot(), which takes two arguments ¹

• data: The data you are going to use
• mapping: Should be wrapped in aes() specifying how each column maps to the coordinate.

# Initialize a ggplot object with 
# data and aesthetic mappings
p <- ggplot(data=iris, 
  mapping = aes(x = Sepal.Length))
p  # Display the base plot (no layers added yet)

# Alternative 
# p <- ggplot(iris, aes(Sepal.Length))
# p

Figure 1: Showing coordinate systems in ggplot2

1. Actually, you can provide them in the later stage instead of here

Layer 2: Plotting data

The general format for plotting items will be

geom_XXX(mapping=aes())

In aes(), you can map anything to the figure, common terms could be

• x
• y
• color, or colour
• fill, shape, etc.

These functions are named in the format geom_XXXX:

	Function name
Histogram	geom_histogram()
Box chart	geom_boxplot()
Bar chart	geom_bar()
Scatter chart	geom_point()
Line chart	geom_line()

Layer 2: Example

p+geom_point(aes(y = Petal.Width, color=Species))

Scatter plot for Sepal length and Petal width

p+geom_boxplot()

Box plot for Sepal length

• + operator to stack on the original ggplot()
• Adding additional mapping in the geom_XXX()
• Overwriting mapping in the ggplot()

More examples

ggplot(data=iris)+
  geom_histogram(aes(x=Petal.Length, color=Species))

Histogram for Petal length

ggplot(data=iris)+
  geom_bar(aes(x=Species,fill=Species))

Histogram for Petal length

Some notes ¹: color affects the border color (dots/lines), fill affects the rectangle color

1. If you find the figure behave not as expected. The best choice is always look up the documentation, or ask AI.

Layer 3: Legends and axis

Scales are functions indicate how each variable behave and presents on the graph. The function usually in the format scale_(AES)_(datatype):

	Discrete	Continuous
X	scale_x_discrete()	scale_x_continuous()
Y	scale_y_discrete()	scale_y_continuous()
color	scale_color_discrete()	scale_color_continuous()

Most of the time you only need to handle the presentation in

Argument name	Axis	Legend
name	Label	Title
breaks	Ticks	Key
labels	Tick label	Key Label

Example

# Create a scatter plot with customized axis and color legend labels
   p +
  geom_point(aes(y = Petal.Width, color=Species)) +  # Add points layer for data representation
  scale_x_continuous(name="Sepal Length") +  # Customize x-axis label
  scale_y_continuous(name="Petal Width") + # Customize y-axis label
  scale_colour_discrete(name="Species name")  # Customize color legend title

Figure 2: Scatter plot of MPG vs cylinders with customized axis and legend labels.

Save figure

Add the function ggsave("FILE_PATH+FILE_NAME.png") after your ggplot instructions:

# Saving plots as figure_theme_grey.png at the folder figure
ggsave("figure_1.png",p)

• It supports most of the common image formats like pdf, jpeg, png

Saved figure location

Final notes

• There are far more things you can do in ggplot2
• Try to look up on the reference (or even use AI) when you need it.

Bibliography notes

Mondrian, Piet. 1935. Composition (No. 1) Gray-Red. Oil on canvas. the Art Institute of Chicago.

Wickham, Hadley. 2010. “A Layered Grammar of Graphics.” Journal of Computational and Graphical Statistics 19 (1): 3–28.

———. 2016. Ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. https://ggplot2.tidyverse.org.