Visualization with R’s tidyverse

ggplot2

Author

Wilfried Cools & Lara Stas

Published

December 27, 2023

SQUARE consultants

square.research.vub.be

Compiled on R 4.3.1

What-Why-Who

This site aims to introduce researchers to visualization in R with the ggplot2 package of the tidyverse ecosystem.

Our target audience is primarily the research community at VUB / UZ Brussel, those who have some basic experience in R and want to know more.

We invite you to help improve this document by sending us feedback: wilfried.cools@vub.be

Key Message

Data visualization can focus on data and/or its summarizing statistics
- to convince yourself, understand data and statistical results
- to convince others, show results to support inference
Data visualization is inherent to data analysis
- more than simply communicating the results
- no -fit’s all data visualization-
- flexible use of data visualization
  - supports more informative and complete visualizations
  - elicits better data exploration and modeling
Data visualization is best done with coding
- maintain structure and transparency, to support reproducibility
Data visualization is easier and more intuitive with tidy data
- tidy data: meaning appropriately mapped into structure
  - each row an observation as research unit,
  - each column a variable as property,
  - each cell a particular value, linking row to column
  - note: data can be split into multiple tables (relational data).
- aim for tidy data registration (avoid tedious manipulations)
Workflow (Hadley Wickham):

R’s tidyverse package: ggplot2

Current focus on ggplot2 the visualization package in the tidyverse eco-system (Hadley Wickham etal.)
Different visualization packages exist in R.
- base R
- grid
- trellis/lattice graphics
- ggplot2 (tidyverse) & ggvis
ggplot2 the current default
- build on idea of Grammar of Graphics (Leland Wilkinson)
  - largely consistent
  - well appreciated defaults
  - easy and intuitive to build (if you get it)
  - without loosing much flexibility
- explicitly links to tidy data
- note: requires extensions for 3D plotting and interactive graphics
- note: does not allow for multiple Y-axes (combination of axes)
ggplot2 part of the tidyverse ecosystem includes:
- dplyr for manipulating data frames [check Data Manipulation]
- tidyr for tidying data [check Data Representation]
- stringr for dealing with texts
- forcats for dealing with factors
- …
Convenient cheat sheets at https://rstudio.com/resources/cheatsheets/.

A quick illustration with ggplot2

toy dataset

The infamous iris data are used
- observe it’s structure with str( ) and first 6 observations head( ) function.
- note: available data with data( )

Have a tidyverse look at the data with glimpse( )

glimpse(iris)

Rows: 150
Columns: 5
$ Sepal.Length <dbl> 5.1, 4.9, 4.7, 4.6, 5.0, 5.4, 4.6, 5.0, 4.4, 4.9, 5.4, 4.…
$ Sepal.Width  <dbl> 3.5, 3.0, 3.2, 3.1, 3.6, 3.9, 3.4, 3.4, 2.9, 3.1, 3.7, 3.…
$ Petal.Length <dbl> 1.4, 1.4, 1.3, 1.5, 1.4, 1.7, 1.4, 1.5, 1.4, 1.5, 1.5, 1.…
$ Petal.Width  <dbl> 0.2, 0.2, 0.2, 0.2, 0.2, 0.4, 0.3, 0.2, 0.2, 0.1, 0.2, 0.…
$ Species      <fct> setosa, setosa, setosa, setosa, setosa, setosa, setosa, s…

Have a tidyverse look at the data with slice_head( )

iris %>% slice_head(n=6)

  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

toy illustration

Make a scatterplot, boxplot, and histogram

p1 <- ggplot(data=iris,
    aes(y=Petal.Width,
        x=Petal.Length,
        col=Species)) + 
    geom_point()

With the iris data
- link dimensions y anx x to Petal columns
- link color to column Species

p1

p2 <- ggplot(data=iris,
    aes(y=Sepal.Length,
        x=Species,
        col=Species)
        ) + 
    geom_boxplot()

x-axis is now Species
Boxplots show a distribution of values for each Species

p2

p3 <- ggplot(data=iris,
    aes(x=Sepal.Width)) + 
    geom_histogram()

x-axis is now Sepal.Width
A histogram shows the full distribution
A warning highlights default use of bin
- Check bins=10

p3

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

ggsave( ) saves the last made plot

ggsave('plotname.png',width=12,height=6)

Layered building blocks

ggplot philosophy: -gg- Grammar of Graphics (Leland Wilkinson)
- build visualization like making a sentence
  - specify building blocks independently
  - combine blocks to create graphical display
- example of layered building blocks:
  - ggplot( data=iris,
    aes(y=Petal.Width,x=Petal.Length,col=Species) ) + geom_point() + geom_smooth()
ggplot( ) creates a ggplot object
- data argument to assign the data frame (or tibble)
- aes( ) function to assign variables from the data frame to scales (x-axis, y-axis, color)
geom_point( ) to add a scatterplot
- note: scatterplots require an x and y-axis
- note: x and y-axis are linked to variables in the data
geom_smooth( ) to add smooted averages
- note: smoothed averages require an x and y-axis
- note: x and y-axis are linked to variables in the data
General structure includes functions and arguments
- functions
  - ggplot( ) initialize the ggplot object
  - geom_*( ) visualize geometric objects
  - stat_*( ) visualize statistical objects
    - largely equivalent to geom
  - facet_*( ) conditional visualization
  - theme( ), guides( ), scale_*( ), coord_*( )
- arguments in ggplot( ), geom_*( ) and stat_*( )
  - data specify data (=input)
  - aes( ) specify aesthetic mapping
    - bridging gap input and output
  - ...
Grammar of Graphics sparked further developments
- other packages with -gg- philosophy: ggforce, ggalt, ggpubr, ggraph, tidygraph, GGally, ggcorrplot, ggridges, ....

Dimensionality

Think of variables to show as dimensions
- combine those variables that show your story
aes( ) links a variable to a dimension
- x: categorical or continuous x-axis
- y: categorical or continuous y-axis
- color:
  - categorical: set of colors
  - continuous: shades of colors within range
- shape: categorical (limited number)
- linewidth: categorical or continuous (dedicated to lines)
- linetype: categorical (dedicated to lines)
- …
Combinations of variables can be linked to one dimension
Facets: categorical (panels)
- a panel for each (combination of) value(s)
Rethink your visualization if this is not sufficient
- use dimensions aimed at bringing focus

Visualization essentials

part 1: how to make a visualization
part 2: how to further refine

step by step example

The ggplot object is constructed
- data is linked to the mtcars data
- x and y aesthetic are linked to it’s variables mpg and disp
- does not visualize !
The internal representation does exist
- ready to extend for visualization
- aesthetics x and y are given their default values
  - mpg and disp from mtcars
- includes a legend and scale for both x and y axis
Any geometric function (object) that at least uses an x and y axis can be added as a layer

ggplot(
    data=mtcars, 
    aes(y=mpg,x=disp)
)

Add a layer with the + sign
- a layer often is geometric function
- a geometric function visualizes a gglot object
geom_point( ): create a scatterplot
- geometric function without arguments

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point()

geom_point() requires data, an x and a y-axis
- if not specified, inherit from ggplot( )
- if specified, ignore ggplot( )
example
- data is not specified inside, but inherited from ggplot( )
  - data is linked to mtcars
- x is not specified inside, but inherited from ggplot( )
  - x linked to variable disp
- y is specified inside within aes( )
  - y linked to variable mpg
get help using ?, ?geom_point

ggplot(data=mtcars, 
    aes(x=disp)
) + 
geom_point(aes(y=mpg))

For this example:
- specify the dimensions in the constructor
- add the data when creating the scatterplot

a tidyverse solution

ggplot() + geom_point(data=mtcars,aes(x=disp,y=mpg))

Non-essential aesthetics like color can be included
- specify color dependent on data within the aes( )
  - color extracted from variable gear
- note: color is a third dimension
Note: the numerical (continuous) variable gear is assigned a continuous scale of colors
- a legend for continuous variables is by default included

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
        color=gear)
) + 
geom_point()

The type of variable determines how it is visualized
A categorical version of the same variable is visualized differently
- the numerical variable is turned into a categorical one (factor())
- a set of colors is used
- a categorical legend is included

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
    color=factor(gear))
) + geom_point()

Instead of color, use shape
- not every dimension is equally clear
- shapes require categorical data

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
        shape=factor(gear))
) + 
geom_point()

Default behavior can be overwritten
- the categorical color variable (geometric function) is used
- the numerical color variable (constructor) is not considered anymore

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,color=gear)) + 
    geom_point(
        aes(color=factor(gear))
    )

Turn the color into shape, but keep the continuously colored scale as well

a tidyverse solution

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,color=gear)) + 
    geom_point(
        aes(shape=factor(gear))
    )

Dimensions can also be linked to constants (not variables in the data)
Assign a dimension outside the aes( ) function
- aes( ) only serves to link dimensions to variables in the data
- color is assigned a value, independent of the data
- color from the ggplot( ) is overwritten

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,color=gear)
) + 
geom_point(color='#FF6600')

Multiple aesthetics (dimensions) can be specified, in and outside the aes( )
With color dependent on gear
- color dependent on a variable is defined inside of aes( )
- alpha (transparency) is set to .3 (30%), outside of aes( )
  - alpha as a percentage (avoid this, it gives unwanted behavior)
  - alpha related to variables with values between 0 and 1

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point(
    aes(color=factor(gear))
    ,alpha=.3)

Note: with alpha (transparency) inside of aes( ) at .3 (30%)
- aes( ) should not be used to link to constants

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point(
    aes(color=factor(gear),alpha=.3)
    )

Use alpha inside of aes( ) to link it to a variable
- alpha related to variables with values between 0 and 1

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point(
    aes(color=factor(gear),
        alpha=qsec/max(qsec)
    )
)

Multiple geometric functions can be included, layered
- geom_point( ) creates the dots, a scatterplot
- geom_line( ) connects them over the x-axis
Note: using the assignment <- code can be build stepwise

myplot <- ggplot(data=mtcars, 
    aes(y=mpg,x=disp)) + 
    geom_line()
myplot <- myplot + geom_line()
myplot

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)) + 
    geom_line()

ggplot(data=mtcars, aes(y=mpg,x=disp)) + 
    geom_point() + geom_line()

Use geom_path( ) to connect subsequent observations
Note: re-ordering has an effect (arrange( ))

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point() + geom_path()

ggplot(data=mtcars %>% arrange(drat), 
    aes(y=mpg,x=disp)
) + 
geom_point() + geom_path()

Add a line over the x-axis, on top of the path
Make the line orange (#FF6600) to highlight it compared to the path

a tidyverse solution

ggplot(data=mtcars %>% arrange(drat), 
    aes(y=mpg,x=disp)
) + 
geom_point() + geom_path() + geom_line(color="#FF6600")

The type of variables in aes( ) determines how it is visualized
Note, assigning a categorical variable to color groups data (lines/path)

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
    color=factor(gear))
) + geom_point() + geom_path()

Note, changing the variable linked to color changes the grouping

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
    color=factor(cyl))
) + geom_point() + geom_path()

Specifications in ggplot( ) offer default behavior
Specifications inside the geom_*( ) overwrite defaults
- note that the geom_point( ) uses the locally specified color aesthetic
- note that the geom_line( ) uses the default black (size made smaller than default)
Beware: types (discrete/continuous) should agree when overwriting ggplot(aes( ))

ggplot(data=mtcars, 
    aes(y=mpg,x=disp)
) + 
geom_point(
    aes(color=factor(gear))
) + 
geom_line(linewidth=.3)

Geometric functions that add statistics can be included as well
- geom_smooth( ) offers averaging and standard errors
  - local averages are default (loess lines)
  - global conditional averages (method=‘lm’)

myplot <- ggplot(data=mtcars, aes(y=mpg,x=disp,color=factor(gear)))
myplot + geom_point() + geom_smooth()

myplot + geom_point() + geom_smooth(method='lm')

Grouping through aesthetics works like before
- the group argument overwrites grouping by other aesthetics like color
- group=1 groups all observations together
Note: get help ?geom_smooth for more details

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,
        color=factor(gear),
        group=1)
) + geom_point() + 
geom_smooth(method='lm',color="#FF6600")

`geom_smooth()` using formula = 'y ~ x'

recap ggplot( )

Function to create a ggplot object, ready for visualization.
- constructor, always required
- prepares appropriate internal representation
- can include default data
- can include default aesthetics

recap aes( )

Function to link variables (data) to an aesthetic (dimension)
- used within ggplot( ) or geom_*( )
- requires variables (part of data)
- defines dimensions dependent on variable values
  - x and y axes: positions on the axes
  - color: color
  - …
- possible arguments depent on geom_*( ) if defined
  - shape: symbols in geom_point( )
  - size: size of bullets in geom_point( )
  - linewidth: width of lines in geom_line( )
  - … check the respective help-files
- group argument used inside aes( ) to identify groups of observations
  - allows grouping without assigning an aesthetic
  - use the value 1 to combine all observations into one group
- automatically assigns a default legend
  - relates to aesthetic
  - depends on variable type (nominal, ordinal, continuous)
- note: aesthetics defined outside of aes( ) are independent of the data

recap geom_*( )

Function to turn an internal ggplot representation into a visualization
- different geoms to create different visualizations
- different geoms use
  - required aesthetics
    - scatterplots require x and y axis
    - histogram requires x axis
  - optional aesthetics (eg., color)
- geom-specific arguments overwrite those inherited from ggplot( )
  - can include aes( )
  - can include data argument
  - useful to add or change aesthetics

Visualization extras

In addition to the basics, there is much more
- stat layers
- scales layers
- facet layer
- theme layer
- coord layer

geom and stat layers

Geometric functions and statistical transformation
- each add layers
- geom_* and stat_* are largely equivalent

A geom_*( ) has a default stat argument
- geom_smooth(stat="smooth")
A stat_*( ) has a default geom argument
- stat_smooth(geom="smooth")
Same result
- + geom_point(stat='summary',fun.y='mean', shape=13,size=16)
- + stat_summary(geom='point',fun.y='mean', shape=13,size=16)
In most cases stick to geom_*( ), occasionally not possible: eg., stat_ecdf( )
- specific stat_*() function highly dedicated

ggplot(data=mtcars, 
    aes(y=mpg,x=factor(carb),
    color=factor(gear))
) + 
geom_point() + 
stat_summary(geom='point',fun.y='mean',shape=13,size=16)

Warning: The `fun.y` argument of `stat_summary()` is deprecated as of ggplot2 3.3.0.
ℹ Please use the `fun` argument instead.

ggplot(mtcars, 
    aes(mpg)
) + stat_ecdf(
    aes(color=factor(cyl)),
    geom = "step"
)

A layer is a function with arguments
- data (see before)
- mapping (aesthetic), defined by the aes( ) (see before)
- geom (eg., point or smooth)
- stat (eg., identity or smooth)
- position (eg., identity)
Position adjustment with position argument
- identity, typically the default
- jitter convenient for points and lines (random perturbation)
- stack, fill and dodge convenient for bars (on top or next to)

myplot <- ggplot(data=mtcars,
    aes(x=disp,
    fill=factor(gear))
) 
myplot + geom_histogram(binwidth=200, 
    position = position_dodge(width=50),
    alpha=.8)

myplot + geom_histogram(binwidth=200, 
    position = position_stack(),
    alpha=.8,
    col='black'
) + theme(legend.position='none')

ggplot(data=mtcars,
    aes(y=mpg,x=drat,color=carb)
) + geom_jitter()

Generated variables are predefined transformations
- start and end with ..
- exist for some geom_*( ) and stat_*( )
- example: a histogram uses a ..density.. to avoid counts

myplot + geom_histogram(
    aes(y=..density..),
    binwidth=200, 
    position = position_stack(),
    alpha=.8,col='black'
) + theme(legend.position='none')

Warning: The dot-dot notation (`..density..`) was deprecated in ggplot2 3.4.0.
ℹ Please use `after_stat(density)` instead.

scale__( )

Each dimension (aesthetic) has a scale
- serves as a legend
- helps with interpretation
Refer to scale with aesthetic and type
- structure: scale__
- examples:
  - scale_x_continuous to assign continuous scale to x-axis aesthetic
  - scale_x_sqrt to square root transform the x-axis aesthetic
  - scale_color_brewer to assign brewer colors to the discrete color aesthetic
  - scale_color_distiller to assign brewer colors to the continuous color aesthetic
  - scale_fill_gradient to assign colors to the fill aesthetic
  - scale_fill_manual to manually assign colors to the fill aesthetic
- arguments control titles, breaks, labels, limits, … see the help file
- note: guide argument requires either a name or guides() function for additional control
Scale types have impact on the visualization and legend (continuous vs. categorical)

ggplot(data=mtcars,
    aes(x=disp,fill=factor(gear))
) + geom_histogram(
    aes(y=..density..),
    binwidth=200, 
    position = position_stack(),
    alpha=.8,
    col='black'
) + theme(legend.position='none') + 
scale_fill_brewer()

ggplot(data=mtcars,
    aes(y=mpg,x=cyl,
    color=drat)
) + geom_point() + 
scale_color_distiller(palette='Oranges')

facet_*( )

Visualizations can be split for subgroups, facilitating conditional comparisons
- facets can be useful to keep plots simple
- by default, the axes are kept constant for comparison, changeable
- facet_grid( ) uses a grid, facet_wrap( ) keeps filling space
Requires a row and/or column specification
- separate rows and columns by ~, use . if none
- use + to combine multiple row and/or column specification

ggplot(data=mtcars,
    aes(y=mpg,x=drat,color=carb)
) + 
geom_jitter() + 
facet_grid(vs~gear)

ggplot(data=mtcars,
    aes(y=mpg,x=drat,color=carb)
) + 
geom_jitter() + 
facet_grid(.~vs+cyl)

theme( )

Themes offer control, unrelated to data
- themes are layers to (re-)specify theme elements
- elements are numerous (?theme):
  - line, rect, text, title
  - axis: axis.title, axis.text.x.top, axis.ticks.x.bottom, …
  - legend: legend.spacing.y, legend.key.width, legend.justification, …
  - panel: panel.background, panel.grid.major, panel.ontop, …
  - plot: plot.background, plot.caption, plot.tag, plot.margin, …
  - strip: strip.background, strip.text, strip.switch.pad.wrap, …
- elements are controlled with a theme function, eg., element_text( )
  - element_text( ), element_line( ), element_rect( ), margin( ), …
- default themes exist, eg., theme_minimal( ), new can be created
- labs( ) is a simpler way to specify titles and labels

(myplot <- ggplot(mtcars, 
    aes(wt, mpg)
) + 
geom_point() + 
labs(
    title = "Fuel economy declines as weight increases",
    y='miles per gallon')
)

The size of the plot title is changed with the element_text( ), the background with the element_rect( )

myplot + theme(plot.title = element_text(size=rel(1.5)), 
    plot.background=element_rect(fill="#FF6600"))

The inside of the plot is the panel, modifiable too.
- multiple elements can be specified in one theme
- multiple themes can be specified by layers

(myplot <- myplot + 
    geom_point(color='#FF6600',size=3) + 
        theme(panel.background = 
            element_rect(fill = "#003399", 
        colour = "#FF6600")
) + 
theme(
    panel.border = element_rect(
        linetype = "dashed", 
        fill = NA
    ), 
    panel.grid.major = element_line(
        colour = "#FF6600")
    )
)

myplot + theme(
  panel.grid.major.y = element_line(colour = "black"),
  panel.grid.minor.y = element_blank()
)

On the outside are the axes and titles, modifiable too
- the text, ticks and titles are adjusted for color and size
- function element_text( ), element_line( ) and unit( ) are used

(myplot <- myplot + theme(
    axis.line = element_line(
        size = 3, 
        colour = "#FF6600")
    ) +
    theme(axis.text = element_text(
        colour = "#003399", 
        size=12)
    ) + 
    theme(axis.ticks.y = 
        element_line(size = 5)
    ) +
    theme(axis.title.y = 
        element_text(size = rel(1.5))
    )
)

Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
ℹ Please use the `linewidth` argument instead.

myplot + theme(
  axis.ticks.length.y = unit(.25, "cm"),
  axis.ticks.length.x = unit(-.25, "cm"),
  axis.text.x = element_text(
    margin = margin(t = .3, unit = "cm")
    )
)

Scales are represented by legends, modifiable too
- labs can be used to change multiple legend titles
- legends can be positioned and formatted

myplot <- ggplot(mtcars, 
    aes(wt, mpg)) +
    geom_point(
        aes(colour = factor(cyl), 
            shape = factor(vs))
    ) +
    labs(
        x = "Weight (1000 lbs)",
        y = "Fuel economy (mpg)",
        colour = "Cylinders",
        shape = "Transmission"
    )

myplot + theme(legend.position='none')

myplot + theme(
    legend.justification = "right",
    legend.position = "bottom"
    )

myplot + theme(
  legend.position = c(.95, .95),
  legend.justification = c("right", "top"),
  legend.box.just = "right",
  legend.margin = margin(6, 6, 6, 6)
)

Keys inside legends are modifiable too
- labs can be used to change multiple legend titles
- legends can be positioned and formatted, for key, text and title

myplot + theme(
    legend.key = element_rect(
        fill = "#bbbbbb", 
        colour = "#003399")
    ) +
    theme(legend.text = element_text(
        size = 14, 
        colour = "#003399")
    ) +
    theme(legend.title = element_text(
        face = "bold")
    )

Themes also work on facets, at which strips are defined

myplot <- ggplot(mtcars, 
        aes(wt, mpg)
    ) + 
    geom_point() + 
    facet_wrap(~ cyl)

myplot + theme(strip.background = 
    element_rect(colour = "black", 
    fill = "white")
)

myplot + theme(strip.text.x = 
    element_text(colour = "white", 
    face = "bold")
)

myplot + theme(panel.spacing = 
    unit(1, "lines")
)

coord_*( )

Typically the default cartesian coordinate system is used, coord_cartesian( )
Limits of the axes are best specified within the coord_*( ) function
- works like a zoom
- alternatively, use xlim( ) and ylim( )
  - beware that values outside the boundary are treated as missing

myplot <- ggplot(data=mtcars,
    aes(y=mpg,x=gear,
    color=factor(vs),
    group=vs)
)
myplot + geom_smooth(method='lm')

`geom_smooth()` using formula = 'y ~ x'

myplot + geom_smooth(method='lm') + 
    coord_cartesian(xlim=c(2.5,4.5))

`geom_smooth()` using formula = 'y ~ x'

myplot + geom_smooth(method='lm') + 
    xlim(2.5,4.5)

`geom_smooth()` using formula = 'y ~ x'

Warning: Removed 5 rows containing non-finite values (`stat_smooth()`).

Within the cartesian family alternatives exist
- coord_flip( ) switches x and y-axis
- coord_fixed( ) sets the ratio for x and y values
  - eg., .1 means 1 unit on x is 10 on y
Alternatives exist, eg., coord_polar( ), coord_trans( ), and various map related functions

myplot + geom_smooth(method='lm') + 
    coord_flip()

`geom_smooth()` using formula = 'y ~ x'

myplot + geom_smooth(method='lm') + 
    coord_fixed(.1)

`geom_smooth()` using formula = 'y ~ x'

myplot + geom_smooth(method='lm') + 
    coord_polar()

`geom_smooth()` using formula = 'y ~ x'

summary

ggplot creates a ggplot object, required for visualization
aes( ) combines aesthetics and link data to scales, and group them
geom_*( ) are geom functions that visualize a ggplot object
stat_*( ) are stat functions that also visualize a ggplot object
scale_*_*( ) helps fine-tuning visualized dimensions
facet_grid( ) or facet_wrap( ) split dimensions over panels (faceting)
coords( ) re-specify the coordinate system, and helps zooming in
theme( ) re-specifies data independent characteristics

Examples to go into detail

Through various examples, the cheat sheet is focused upon.
- https://rstudio.com/resources/cheatsheets/

one-variable

Various visualizations address one particular variable
- mostly continuous but possibly also discrete
- continuous variables are typically ‘binned’
Note the frequency polygon, and the histogram.

ggplot(data=mtcars,aes(mpg)) + 
geom_freqpoly(binwidth=2.5)

ggplot(data=mtcars,aes(mpg)) + 
geom_histogram(binwidth=2.5)

Using the geom_area( ), binning must be explicit as stat argument.

ggplot(data=mtcars,aes(mpg)) + 
    geom_area(stat='bin',binwidth=2.5)

Continuous variables can also be shown continuously, without binning
Different types of densities are shown

ggplot(data=mtcars,aes(mpg)) + geom_density() + 
    geom_density(kernel='triangular',color="#003399") + geom_density(kernel='optcosine',color="#FF6600")

The geom_qq( ) requires a sample argument (instead of x)
- positions are determined by their value

ggplot(data=mtcars,
    aes(sample=mpg)) + 
    geom_qq()

The bar-plot with geom_bar( ) is similar to the histogram
- shows the actual values instead of a count per bin

ggplot(data=mtcars,
    aes(factor(cyl))) + 
    geom_bar(fill='#FF6600',color='#003399')

exercises on one variable visualizations

Make use of the mpg dataset again
Make a histogram for the continuously scaled displ variable

an R solution

ggplot(data=mpg,aes(x=displ)) + 
    geom_histogram()

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

Adjust the binwidth to .5
Have a glimpse at the data, what data-types are included ?

a tidyverse solution

ggplot(data=mpg,aes(x=displ)) + 
    geom_histogram(binwidth=.5)

Add a frequency polynomial on top (freqpoly)

a tidyverse solution

ggplot(data=mpg,aes(x=displ)) + 
    geom_histogram(binwidth=.5) + 
    geom_freqpoly()

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

Notice what happens if the same binwidth is used for the frequency polynomial.

a tidyverse solution

ggplot(data=mpg,aes(x=displ)) + 
    geom_histogram(binwidth=.5) + 
    geom_freqpoly(binwidth=.5)

Make a barplot for the discretely scaled class variable

a tidyverse solution

ggplot(data=mpg,aes(x=class)) + 
    geom_bar()

Group the data by filling in colors dependent on the drv variable

a tidyverse solution

ggplot(data=mpg,aes(x=class)) + 
    geom_bar(aes(fill=drv))

Turn the bars next to one-another
Reduce their width to .5 to increase space between bars

a tidyverse solution

ggplot(data=mpg,aes(x=class)) + 
    geom_bar(aes(fill=drv),
    position='dodge',width=.5)

two-variables

Various visualizations address the relation between two variables, whether discrete and/or continuous.
Especially for categorical data data could obscure other data.
- deal with this using the position argument or with the geom_jitter( )
- avoid combining both geom_point( ) and geom_jitter( ) as it would draw points each time

ggplot(data=mtcars,aes(cyl,gear)) + 
geom_point()

ggplot(data=mtcars,aes(cyl,gear)) + 
geom_jitter(width=.2,height=.2)

ggplot(data=mtcars,aes(cyl,gear)) + 
geom_point(color='#FF6600') + 
geom_jitter(width=.2,height=.2)

The smooth function has been shown above
The confidence band with the middle 50% (quantiles .25 and .75) is used
A rug at the axes captures the one dimensional distribution.
Instead of bullet indicators, the row names (or any other set of labels) can be used
- use the label argument (within the aes( ) when related to data)
- some jitter reduces overlap

ggplot(data=mtcars, 
    aes(y=mpg,x=disp,color=factor(gear))) + 
    geom_point() + 
    geom_smooth(method='lm',se=FALSE) + 
    geom_rug() + 
    geom_quantile(quantiles=c(.25,.75),linetype=2)

`geom_smooth()` using formula = 'y ~ x'
Smoothing formula not specified. Using: y ~ x
Smoothing formula not specified. Using: y ~ x
Smoothing formula not specified. Using: y ~ x

ggplot(mtcars, aes(wt, mpg)) + 
geom_text(
    aes(label=(rownames(mtcars))),
    size=2,
    position=position_jitter(width = .2,height=3, seed=256)
)

ggplot(data=mtcars,aes(cyl,gear)) + 
geom_point(color='#FF6600') + 
geom_jitter(width=.2,height=.2)

Bars can be obtained with geom_col( ), or with geom_bar(stat='identity')
- use of ‘identity’ causes the height to depend on the numbers in the data
- Note: these numbers are summed if not unique, be careful.
A count( ) extracts the frequency of the specified grouping

(freq_by_group <- mtcars %>% group_by(gear,vs) %>% count())

# A tibble: 6 × 3
# Groups:   gear, vs [6]
   gear    vs     n
  <dbl> <dbl> <int>
1     3     0    12
2     3     1     3
3     4     0     2
4     4     1    10
5     5     0     4
6     5     1     1

ggplot(freq_by_group, 
    aes(fill=factor(vs), y=n, x=gear)) + 
geom_bar(position="stack", stat="identity")

ggplot(freq_by_group, 
    aes(fill=factor(vs), y=n, x=gear)) + 
geom_col(position="dodge")

ggplot(mtcars, 
    aes(fill=factor(vs), y=mpg, x=gear)) + 
geom_col(position="dodge") + 
labs(y='sums all mpg values')

A boxplot nicely summarizes continuous data, possibly for different groups

ggplot(data=mtcars,
    aes(y=mpg,x=factor(cyl))) + 
geom_boxplot()

ggplot(data=mtcars,
    aes(y=mpg,x=factor(cyl))) + 
    geom_boxplot(width=.25,alpha=.2,
        aes(fill=factor(cyl))
    ) + 
    geom_jitter(width=.05)

exercises on two variable visualizations

Make a scatterplot for the continuously scaled hwy on cty, and color by cyl

an R solution

ggplot(data=mpg,aes(y=hwy,x=cty,color=cyl)) + 
    geom_point()

Jitter the data, so it shows if data points obscure one-another
Make sure that cyl is categorical

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=cty,color=factor(cyl))) + 
geom_jitter() + 
geom_smooth(method='lm')

`geom_smooth()` using formula = 'y ~ x'

Add a conditional average with geom_smooth( ), use the lm method

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=cty,color=class)) + 
geom_jitter(aes(shape=factor(cyl))) + 
geom_smooth(method='lm')

`geom_smooth()` using formula = 'y ~ x'

Add a shape dependent on class, notice the restriction on the number of shapes

a tidyverse solution

ggplot(data=mpg,aes(y=hwy,x=cty)) + 
    geom_jitter(
        aes(shape=factor(cyl))
    ) + 
    geom_smooth(method='lm',aes(color=class))

`geom_smooth()` using formula = 'y ~ x'

Switch the color and shapes around (color can have more than 6 if it is really necessary)

a tidyverse solution

ggplot(data=mpg,aes(y=hwy,x=cty)) + 
    geom_jitter(aes(shape=factor(cyl))) + geom_smooth(method='lm',aes(color=class))

`geom_smooth()` using formula = 'y ~ x'

Make sure the symbols do not differ by color (all black), only shape, but keep the regression lines

a tidyverse solution

ggplot(data=mpg,aes(y=hwy,x=cty)) + 
    geom_jitter(aes(shape=factor(cyl))) + geom_smooth(method='lm',aes(color=class))

`geom_smooth()` using formula = 'y ~ x'

Show boxplots for the hwy for each cyl

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(cyl))
) + 
geom_boxplot()

Add the actual data points and make sure they do not obscure each other

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(cyl))
) + 
geom_boxplot() + 
geom_jitter()

Give a color to the observations dependent on class

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(cyl))
) + 
geom_boxplot() + 
geom_jitter(aes(color=class))

Make a bar chart that adds all hwy values in each class group

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(class))
) + 
geom_col()

Use a coloring of the bars to signal the relative contribution of all cyl categories

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(class),
    fill=factor(cyl))
) + 
geom_col()

Flip the coordinates x and y axis

a tidyverse solution

ggplot(data=mpg,
    aes(y=hwy,x=factor(class),
    fill=factor(cyl))
) + 
geom_col() + 
coord_flip()

intervals

Specialized functions facilitate visualization of errors / confidence intervals
Standard errors or other intervals can be visualized along with fitted values

(tmp <- tribble(
~set,~fit,~se,
1,3,.2,
2,2,.3,
3,2,.4))

# A tibble: 3 × 3
    set   fit    se
  <dbl> <dbl> <dbl>
1     1     3   0.2
2     2     2   0.3
3     3     2   0.4

(myplot <- ggplot(data=tmp,
    aes(y=fit,x=set)))

myplot + geom_errorbar(
    aes(ymax=fit+2*se,ymin=fit-2*se)
)

myplot + geom_errorbarh(
    aes(xmax=fit+2*se,xmin=fit-2*se)
)

myplot + geom_crossbar(
    aes(ymax=fit+2*se,ymin=fit-2*se)
)

myplot + geom_pointrange(
    aes(ymax=fit+2*se,ymin=fit-2*se)
)

third variable implied

Frequencies and densities can be obtained for two variables
With geom_bin2d( ) the frequency of combinations is obtained
- in this case for the factors cyl and gear
Continuous variables can be used with binning
Contours can be obtained with geom_density2d( )

ggplot(data=mtcars,
    aes(y=factor(cyl),x=factor(gear))
) + 
geom_bin2d()

ggplot(data=mtcars,
    aes(y=mpg,x=carb)
) + geom_density2d(
    aes(colour = factor(gear))
)

three variables

A z dimension is possible, while typically other aesthetics are used
A heatmap is the most obvious use
- to show for example correlations between many variables with colors instead of values
- a small example is used instead
  - notice the argument for tile is a fill
  - z is the argument for the contour

tmp <- expand.grid(set1=1:10,set2=1:10); 
set.seed(123); 
tmp$score <- runif(100,0,1)
head(tmp)

  set1 set2     score
1    1    1 0.2875775
2    2    1 0.7883051
3    3    1 0.4089769
4    4    1 0.8830174
5    5    1 0.9404673
6    6    1 0.0455565

ggplot(data=tmp,
    aes(y=set2,x=set1)
) + 
geom_tile(aes(fill=score))

ggplot(data=tmp,
    aes(y=set2,x=set1)
) + 
geom_contour(aes(z=score))

primitives

Primitives are the basic building blocks
- several primitives exist: point( ), path( ), polygon( ), segment( ), ribbon( ), rect( ), text( ), blank( )
- only geom_point( ) is used very often
- geom_ribbon( ) can be interesting for showing intervals
- typically useful for fine-tuning only

The polygon( ) is used, with coordinates created in a separate datafile.
The same is done with tiles( )
It is possible to create the datafile within the function.
First a blank plot is drawn, then the rest is added.

blank_plot <- ggplot(mtcars,
    aes(y=mpg,x=cyl)) + 
    geom_blank()

polygon_coordinate_file <- tribble(
~x,~y,
6,17,
6,10,
5,10,
8,21)

tile_coordinate_file <- tribble(
~x,~y,~w,
5,20,2,
6,30,5,
7,15,2)

(updated_plot <- blank_plot + 
    geom_polygon(data=polygon_coordinate_file,
    aes(x=x,y=y),alpha=.3,color="#FF6600"))

(updated_plot <- updated_plot + 
    geom_tile(data=tile_coordinate_file,
        aes(x=x,y=y,width=2),alpha=.3,
        color="#003399"))

updated_plot + geom_segment(data=
    data.frame(x=5,y=15,xend=7,yend=20),
    aes(x=x,y=y,xend=xend,yend=yend),
    size=5,color="#FF6600")

Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.

exercises on primitives

The mpg dataset is part of tidyverse’s ggplot !

an R solution

data(mpg)

Have a glimpse at the data, what data-types are included ?

a tidyverse solution

glimpse(mpg)

Rows: 234
Columns: 11
$ manufacturer <chr> "audi", "audi", "audi", "audi", "audi", "audi", "audi", "…
$ model        <chr> "a4", "a4", "a4", "a4", "a4", "a4", "a4", "a4 quattro", "…
$ displ        <dbl> 1.8, 1.8, 2.0, 2.0, 2.8, 2.8, 3.1, 1.8, 1.8, 2.0, 2.0, 2.…
$ year         <int> 1999, 1999, 2008, 2008, 1999, 1999, 2008, 1999, 1999, 200…
$ cyl          <int> 4, 4, 4, 4, 6, 6, 6, 4, 4, 4, 4, 6, 6, 6, 6, 6, 6, 8, 8, …
$ trans        <chr> "auto(l5)", "manual(m5)", "manual(m6)", "auto(av)", "auto…
$ drv          <chr> "f", "f", "f", "f", "f", "f", "f", "4", "4", "4", "4", "4…
$ cty          <int> 18, 21, 20, 21, 16, 18, 18, 18, 16, 20, 19, 15, 17, 17, 1…
$ hwy          <int> 29, 29, 31, 30, 26, 26, 27, 26, 25, 28, 27, 25, 25, 25, 2…
$ fl           <chr> "p", "p", "p", "p", "p", "p", "p", "p", "p", "p", "p", "p…
$ class        <chr> "compact", "compact", "compact", "compact", "compact", "c…

Make a scatterplot for cty per hwy, and assign it to the object name ‘current’

a tidyverse solution

current <- ggplot(data=mpg,aes(y=hwy,x=cty)) + geom_point()

Draw 2 rectangles with geom_rect( ),
- one with corners 10,15 and 20,25,
- one shifted with 5 both dimensions,
- using 2 additional data frames (tibbles) to store the coordinates.

# tmp <- data.frame(
    # c1=c(10,15),
    # c2=c(15,20),
    # c3=c(25,30),
    # c4=c(35,40))
my_2_sets_of_coordinates <- tribble(~xi,~yi,~xa,~ya,
10,15,25,35,
15,20,30,40
)

Plot the coordinates on top of the scatterplot
Note: the argument inherit.aes=FALSE avoids inheriting aesthetics

a tidyverse solution

current + geom_rect(data=my_2_sets_of_coordinates,
    aes(xmin=xi,ymin=yi,xmax=xa,ymax=ya),
    inherit.aes=FALSE,alpha=.3)

Add a different color to each separately, use ‘#FF6600’ and ‘#003399’

a tidyverse solution

current + 
    geom_rect(data=my_2_sets_of_coordinates[1,],
        aes(xmin=xi,ymin=yi,xmax=xa,ymax=ya),
        inherit.aes=FALSE,alpha=.3,
        fill="#003399") + 
    geom_rect(data=my_2_sets_of_coordinates[2,],
        aes(xmin=xi,ymin=yi,xmax=xa,ymax=ya),
        inherit.aes=FALSE,alpha=.3,
        fill="#FF6600")

Beyond ggplot2: -gg- packages

Various extensions have been developed, each showing the same basic structure but typically addressing particular types of visualizations
Check https://exts.ggplot2.tidyverse.org/gallery/

library(ggfortify)  
autoplot(scale(mtcars))

Scale for y is already present.
Adding another scale for y, which will replace the existing scale.

library(ggforce)
ggplot(iris, aes(Sepal.Length, Sepal.Width)) + 
  geom_voronoi_tile(aes(fill = Species, group = -1L)) + 
  geom_voronoi_segment() +
  geom_point()

Warning: `stat_voronoi_tile()` is dropping duplicated points

Warning: `stat_voronoi_segment()` is dropping duplicated points

library(gganimate)
library(gapminder)
ggplot(gapminder, aes(gdpPercap, lifeExp, size = pop, colour = country)) +
  geom_point(alpha = 0.7, show.legend = FALSE) +
  scale_colour_manual(values = country_colors) +
  scale_size(range = c(2, 12)) +
  scale_x_log10() +
  facet_wrap(~continent) +
  labs(title = 'Year: {frame_time}', x = 'GDP per capita', y = 'life expectancy') +
  transition_time(year) +
  ease_aes('linear')

Last remarks

Current page provides a primer on visualization, which is a key tool for understanding data and statistics required within the process of data analyses.

Get ahead at https://ggplot2-book.org/

It is strongly advised to play with the techniques discussed above to get some proficiency in using it, as it would add significantly to the flexibility of whatever you want to further do with your data.

Other tidyverse package exist, and within the same framework many more are being developed. The consistency within the tidyverse ecosystem should give you a push though, to study the other packages yourself when of interest.

Base R still is a proper alternative to the tidyverse ecosystem, so be aware that others may do things differently.

R’s tidyverse package: ggplot2

A quick illustration with ggplot2

toy dataset

toy illustration

Layered building blocks

Dimensionality

Visualization essentials

step by step example

recap ggplot( )

recap aes( )

recap geom_*( )

Visualization extras

geom and stat layers

scale_*_*( )

facet_*( )

theme( )

coord_*( )

summary

Examples to go into detail

one-variable

exercises on one variable visualizations

two-variables

exercises on two variable visualizations

intervals

third variable implied

three variables

primitives

exercises on primitives

Beyond ggplot2: -gg- packages

Last remarks

scale__( )