PSYC40940: Grammar of Graphics

From Principles To Practice

~~Week 2: Principles of data visualisation~~
Week 3: Grammar of Graphics; aesthetics and attributes
Week 4: Major visualisation tools
Week 5: Customising visualisations (scales, themes, and labels)

What You Should Take Away

By the end of the session, you should be able to:

read a ggplot2 plot as a set of layers
explain the difference between aesthetics and attributes
choose simple aesthetic mappings for continuous and categorical variables
avoid overloaded mappings that make plots harder to read

Download the Week 3 files from NOW and move them into your RStudio Project folder.

There are two exercise files for today:

01_building_ggplot_layers.Rmd
02_aesthetics_vs_attributes.Rmd

Building Plots With Grammar

What Parts Make Up This Plot?

Before we name the grammar, look at the plot and identify what it is made from.

What do you notice about the data, visual marks, colours, lines, labels, and layout?

Core Components

Component	Example	Description
Data	`d_spellname`	The dataset being plotted
Aesthetics (aes)	`x = dur, y = rt`	How data variables map to visual properties
Geometries (geom)	`geom_point()`	The type of plot element (points, bars, lines)
Statistics (stat)	`stat_smooth()`	Transformations of data (e.g., regression lines)
Scales	`scale_x_log10()`	How data values are converted to aesthetics
Facets	`facet_wrap(~group)`	Splitting data into small multiples
Coordinate system	`coord_polar()`	The space where data are drawn (Cartesian, polar)
Theme	`theme_minimal()`	Non-data display elements (fonts, text, grids, etc.)

Why `ggplot2` Uses A Grammar

Wilkinson (2005)

ggplot2 is an R package for creating data visualisations.
The “gg” stands for Grammar of Graphics (Wilkinson, 2005).
A grammar means plots are built from reusable parts, rather than memorised as separate plot types.
This helps you start simple and add complexity only when it helps the reader.
See also Wickham (2016), Wickham (2010).

What The Grammar Describes

Wilkinson (2005)

A statistical graphic connects three main things:

data: the variables we want to show
aesthetics: how variables become visual properties such as position, colour, shape, or size
geometries: the visible marks, such as points, lines, or bars

Other layers can then summarise, split, transform, label, or style the plot.

Grammar As Building Blocks

Wilkinson (2005)

Think about language: grammar tells us how words can be combined into sentences.

For plots, the grammar tells us how data, mappings, and visual marks can be combined.

The practical skill for today is to read a plot as layers: What data are used? What is mapped? What is drawn? What else has been added?

The Three Required Parts

data: the data you want to visualise indicated as ggplot(data = ...)
aesthetics: mapping of data to graphic properties (axes, size, colour) indicated as mapping = aes()
geometries: visual elements encoding the data indicated as geom_...()

Required Parts: Data And Aesthetics

ggplot(d_spellname, aes(x = dur, y = rt))

Add A Geometry: Points

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point()

Change The Geometry: Quantiles

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_quantile()

Change The Geometry: Rug Marks

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_rug()

Combine Multiple Geometries

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  geom_quantile() +
  geom_rug()

Layers You Can Add Later

facets: dividing data into subplots
statistics: statistical summaries
coordinates: how the plotting space is arranged
theme: visual properties not related to the data (font, background)

Start With Data, Aesthetics, And Geometry

data
aesthetics
geometries

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point()

Add Facets To Split The Plot

Add A Statistical Summary

data
aesthetics
geometries
facets
statistics

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  stat_smooth(method = "lm", se = FALSE)

Change The Coordinate System

data
aesthetics
geometries
facets
statistics
coordinates

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  coord_trans(x = "log", y = "log")

Flip The Coordinates

data
aesthetics
geometries
facets
statistics
coordinates

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  coord_flip()

Add A Theme

data
aesthetics
geometries
facets
statistics
coordinates
theme

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  theme_dark()

Adjust One Theme Element

data
aesthetics
geometries
facets
statistics
coordinates
theme

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  theme(panel.background = element_blank())

Exercise 1: Build A Plot Layer By Layer

Complete RMarkdown document 01_building_ggplot_layers.Rmd.

This exercise practises reading a plot as a sequence of ggplot2 layers.

Read The Code As Layers

ggplot(data = d_spellname, 
       mapping = aes(x = dur, 
                     y = rt, 
                     colour = group)) +
  geom_point(size = .75) +
  stat_smooth(method = "lm", se = F) +
  labs(x = "Response duration (in ms)",
       y = "Reaction time (in ms)") +
  theme(legend.position = "right",
       legend.justification = "top",
       legend.direction = "vertical")

Mapping Variables And Setting Styles

Mapping Versus Setting

Aesthetics: visual properties mapped to variables (data-driven) e.g. aes(x = duration, y = time, colour = group) → colour changes with data and usually creates a scale / legend

Attributes: visual properties set manually (constant) e.g. geom_point(colour = "blue", size = 2) → colour is fixed and does not describe a variable

Think:

Aesthetic = mapping (data → visual)

Attribute = setting (visual → fixed property)

Attribute: Set One Colour

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point(colour = "red")

Aesthetic: Map Colour To A Variable

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point(aes(colour = modality))

Mappings Can Affect Later Layers

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point(aes(colour = modality)) +
  stat_smooth(method = "lm")

Layer-Specific Mappings

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point() +
  stat_smooth(aes(colour = modality), method = "lm")

Repeating A Mapping Across Layers

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt)) +
  geom_point(aes(colour = modality)) +
  stat_smooth(aes(colour = modality), method = "lm")

Global Mappings Apply To All Layers

appearance of geometries
e.g. colour, size, shape
attributes take properties
aesthetics take variables

ggplot(d_spellname, aes(x = dur, y = rt, colour = modality)) +
  geom_point() +
  stat_smooth(method = "lm")

Shape Can Encode Groups

ggplot(d_spellname, aes(x = dur, y = rt, shape = modality)) +
  geom_point(size = 2.5, colour = "red")

Colour Can Encode Groups

ggplot(d_spellname, aes(x = dur, y = rt, colour = modality)) +
  geom_point(size = 2.5)

Avoid Overloading A Plot

ggplot(d_spellname, 
       aes(x = dur, y = rt, colour = modality)) +
  geom_point(size = 2.5)

Use one main aesthetic when it answers the question clearly.

Colour, shape, size, and labels all compete for attention.
Add another aesthetic only when it improves comparison or accessibility.
More encodings do not automatically make a plot more informative.

Use Redundant Coding Carefully

ggplot(d_spellname, aes(x = dur, y = rt,
                        colour = modality,
                        shape = modality)) +
  geom_point(size = 2.5)

Sometimes redundant coding is helpful:

colour helps quick grouping
shape can help when printed in greyscale
two groups are still manageable

Avoid this when it creates too many visual groups.

Common Aesthetics To Recognise

some are required by geometries; others are optional
continuous vs discrete variables:
- e.g. shape and label can only be used for categorical values
should be chosen to facilitate comprehension

scatterplot: geom_point()

x, y, shape, colour, fill, size, alpha, stroke, group

barplot: geom_bar()

x, y, colour, fill, linewidth, linetype, alpha, width, group

boxplot: geom_boxplot()

x, y, lower, xlower, upper, xupper, middle, xmiddle, ymin, 
xmin, ymax, xmax, weight, colour, fill, size, alpha, shape, 
linetype, linewidth, width, group

Exercise 2: Aesthetics And Attributes

Complete RMarkdown document 02_aesthetics_vs_attributes.Rmd.

This exercise practises deciding whether a visual property should be mapped to a variable or set to a fixed value.

Variable Type Matters

Before mapping a variable to colour, shape, size, or another aesthetic, ask what kind of variable it is.

Variable type	Examples	Usually works well with
Continuous	age, reaction time, vocabulary score	position, gradients, sometimes size
Categorical	modality, group, condition	position, colour hue, shape, linetype, facets

For today, focus on where aesthetics go in the code. We will use variable type more directly next week when choosing plot types.

Check Your Mappings

If time permits, revisit your plots and ask whether each aesthetic improves the comparison.

Use the two exercise files from today as your main practice for Week 3.

Homework

Identify a dataset you would like to use for the formative assessment: post a description on our Teams channel.

Think about what you would like to visualise about these data:

Combination of variables
Variable types and appropriate aesthetics
Raw data and / or summary statistics
Look for appropriate visualisation tools HERE

References

Andrews, M. (2021). Doing data science in R: An introduction for Social Scientists. SAGE Publications Ltd.

Wickham, H. (2010). A layered grammar of graphics. Journal of Computational and Graphical Statistics, 19(1), 3–28.

Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer.

Wickham, H., & Grolemund, G. (2016). R for data science: Import, tidy, transform, visualize, and model data. O’Reilly Media, Inc.

Wilkinson, L. (2005). The grammar of graphics (2nd ed.). Springer.

From Principles To Practice

What You Should Take Away

Building Plots With Grammar

What Parts Make Up This Plot?

Core Components

Why ggplot2 Uses A Grammar

Wilkinson (2005)

What The Grammar Describes

Wilkinson (2005)

Grammar As Building Blocks

Wilkinson (2005)

The Three Required Parts

Required Parts: Data And Aesthetics

Add A Geometry: Points

Change The Geometry: Quantiles

Change The Geometry: Rug Marks

Combine Multiple Geometries

Layers You Can Add Later

Start With Data, Aesthetics, And Geometry

Add Facets To Split The Plot

Add A Statistical Summary

Change The Coordinate System

Flip The Coordinates

Add A Theme

Adjust One Theme Element

Exercise 1: Build A Plot Layer By Layer

Read The Code As Layers

Mapping Variables And Setting Styles

Mapping Versus Setting

Attribute: Set One Colour

Aesthetic: Map Colour To A Variable

Mappings Can Affect Later Layers

Layer-Specific Mappings

Repeating A Mapping Across Layers

Global Mappings Apply To All Layers

Shape Can Encode Groups

Colour Can Encode Groups

Avoid Overloading A Plot

Use Redundant Coding Carefully

Common Aesthetics To Recognise

Exercise 2: Aesthetics And Attributes

Variable Type Matters

Check Your Mappings

Recommended Reading

Homework

References

Why `ggplot2` Uses A Grammar