The Grammar of Graphics

Dr. Thiyanga S. Talagala
Department of Statistics, Faculty of Applied Sciences
University of Sri Jayewardenepura, Sri Lanka

Packages we need today

library(ggplot2)
library(palmerpenguins)

Pioneers

Concept: Leland Wilkinson

Implementation: Hadley Wickham

Building Blocks of layers with the grammar of graphics

1. Data

2. Aesthetics

3. Geometrics

4. Facets

5. Scales

6. Statistics

7. Coordinates

8. Themes

Data

A data frame or tibble containing data for the plot.

Ingeridients to do the plotting

Step 1: Take Data

data(penguins)

Step 2: Take a canvas to do the plot

ggplot()

Process: drawing the plot

Take your data onto your canvas for plotting

ggplot(data=penguins)

Aesthetics

How variables in the data are mapped to visual properties (aesthetics) of geometries (geoms)

Data + Aesthetics

Geometrics (Geoms)

The visual elements used to plot data

Data + Aesthetics + Geom

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point()

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_smooth()

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_smooth(method="lm")

Facets

Plotting small multiples (Partition the data into smaller “sub plots”, or panels)

Data + Aesthetics + Geom + Facet

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_wrap(~species)

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_wrap(~species, ncol=3)

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_wrap(~species, nrow=3)

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_wrap(~species, nrow=3, strip.position = "left") 

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island)

Scale

cales control the mapping between data values and aesthetics like color, size, and shape. You can customize scales using functions like scale_x_continuous(), scale_fill_manual(), etc., to adjust the appearance of the plot.

Data + Aesthetics + Geom + Facet + Scale

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_colour_manual(values = c("#1b9e77","#d95f02","#7570b3"))

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_colour_manual(values = c("#1b9e77","red","#7570b3"))

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_colour_manual(values = c("Gentoo"="#1b9e77", "Chinstrap"="red", "Adelie" = "#7570b3"))

RColorBrewer::display.brewer.all(type = "div")

RColorBrewer::display.brewer.all(type = "seq")

RColorBrewer::display.brewer.all(type = "qual")

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + facet_grid(species ~ island) + scale_color_brewer(palette = "Dark2")

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + facet_grid(species ~ island) +
scale_color_manual(values = RColorBrewer::brewer.pal(n = 3, name = "Set1"))

Statistics

Visualize statistical measures

Data + Aesthetics + Geom + Facet + Stat

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) +
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 7,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  )

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  )

Coordinates

Controls coordinate system

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_flip()

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_fixed(0.03)

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_fixed(1)

Themes

Themes control non-data elements, including elements like background color, grid lines, axis labels, and titles.

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_cartesian() +
theme(title = element_text("Penguin flipper length vs body mass"),
       axis.title.x =  element_text("Flipper length (mm)"),
       axis.title.y =  element_text("Body mass (g)"),
legend.position = "bottom")

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_cartesian() + theme_bw() + 
theme(title = element_text("Penguin flipper length vs body mass"),
       axis.title.x =  element_text("Flipper length (mm)"),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1),
       axis.title.y =  element_text("Body mass (g)"),
legend.position = "bottom") 

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_cartesian() + 
theme(title = element_text("Penguin flipper length vs body mass"),
       axis.title.x =  element_text("Flipper length (mm)"),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1),
       axis.title.y =  element_text("Body mass (g)"),
legend.position = "bottom") + theme_bw() 

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_cartesian() + 
theme(title = element_text("Penguin flipper length vs body mass"),
       axis.title.x =  element_text("Flipper length (mm)"),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1),
       axis.title.y =  element_text("Body mass (g)"),
legend.position = "bottom") + theme_dark() 

Labs

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
facet_grid(species ~ island) + 
scale_color_brewer(palette = "Dark2") +
stat_summary(
    geom = "point",
    fun.y = "mean",
    col = "black",
    size = 3,
    shape = 24,
    fill = "red",
    alpha = 0.5,
  ) + coord_cartesian() + 
labs(tile = "Penguin flipper length vs body mass",
     x = "Flipper length (mm)",
     y = "Body mass (g)") + 
theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1), legend.position = "bottom") + theme_dark() 

Local Data vs Global Data

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
coord_fixed(ratio=0.02)

I want to mark (200, 4500) on the plot

df <- tibble::tibble(x=200, y=4500)

ggplot(data=penguins, aes(x=flipper_length_mm, y=body_mass_g, col= species)) +
geom_point() + 
geom_point(data=df, aes(x=x, y=y), col="black", size=5) + 
coord_fixed(ratio=0.02)