Data plotting

October 2019

ggplot2

Learning objectives

You will learn to:

Understand the basic grammar of graphics
How it is implemented in ggplot2
- input data, data.frame/tibble
- aesthetics
- geoms
- facets
- themes
Make quick exploratory plots of your multidimensional data.
Know how to find help on ggplot2 when you run into problems.

Introduction

`ggplot2`

stands for grammar of graphics plot v2
Inspired by Leland Wilkinsons work on the grammar of graphics in 2005.

Idea: split a graph into layers

such as axis, curve(s), labels.
3 main elements are required: data, aesthetics, geometry \(\geqslant 1\)

Wickham H. 2007. J. Comp. Graph. Stat. 19:3–28

source: thinkR

Simple example

Wickham 2007

dataset

x	y	shape
25	11	circle
0	0	circle
75	53	square
200	300	square

aesthetics

x = x, y = y, shape = shape

geometric object

dot / point

What if we want to split circles and squares?

Faceting

aka trellis or lattice plots

Split by the shape

Redundancy

Now, dot shapes and facets provide the same information.
We could use the shape for another meaningful variable…

Layers

Data

x	y	shape
25	11	circle
0	0	circle
75	53	square
200	300	square

Motivation for this layered system

Data visualisation is not meant just to be seen but to be read, like written text Alberto Cairo

Geometric objects

geoms define the type of plot which will be drawn

`geom_point()`

`geom_line()`

`geom_bar()`

`geom_boxplot()`

`geom_histogram()`

`geom_density()`

Cheatsheet

Have a look at the cheatsheet or the ggplot2 online documentation to list more possibilities.

The dataset

Convert iris as a `tibble`

iris <- as_tibble(iris)
iris

# A tibble: 150 x 5
   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
          <dbl>       <dbl>        <dbl>       <dbl> <fct>  
 1          5.1         3.5          1.4         0.2 setosa 
 2          4.9         3            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           3.6          1.4         0.2 setosa 
 6          5.4         3.9          1.7         0.4 setosa 
 7          4.6         3.4          1.4         0.3 setosa 
 8          5           3.4          1.5         0.2 setosa 
 9          4.4         2.9          1.4         0.2 setosa 
10          4.9         3.1          1.5         0.1 setosa 
# … with 140 more rows

Tip

saving the data frame as a tibble enables the smart tibble printing and avoids to list all 150 rows

Your first plot

Draw your first plot

ggplot(data = iris) +
  geom_point(mapping = aes(x = Petal.Width, y = Petal.Length))

Layers and pipes

Warning

ggplot2 introduces a break in the workflow from %>% to +

`ggplot1`

ggplot1 was released in 2005 until 2008 by Hadley Wickham.

If the pipe ( %>% in 2014) had been invented before, ggplot2 would have never existed Hadley Wickham

`ggplot1`: original syntax

# devtools::install_github("hadley/ggplot1")
library(ggplot1)
p <- ggplot(mtcars, list(x = mpg, y = wt))
# need temp p object to avoid too many ()'s
scbrewer(ggpoint(p, list(colour = gear)))

`ggplot1` with the pipe

library(ggplot1)
mtcars %>% 
  ggplot(list(x = mpg, y = wt)) %>% 
  ggpoint(list(colour = gear)) %>% 
  scbrewer()

`ggplot2`

library(ggplot2)
mtcars %>% 
  ggplot(aes(x = mpg, y = wt)) + 
  geom_point(aes(colour = as.factor(gear))) +
  scale_colour_brewer("gear", type = "qual")

Mapping aesthetics

definitions

aesthetics map the columns of a tibble to the variable each ggplot2 geom is expecting.
geom_point() for example requires at least the x and y coordinates to draw each point.

In our example we need to tell geom_point() which columns should be used as x and y

ggplot(iris) +
  geom_point(aes(x = Petal.Width, y = Petal.Length))

aesthetics requirements

each geom has specific requirement depending on its input
- univariate, one x like histogram or density
- bivariate, x and y like scatterplot
and requirements for continuous or discrete variable
- geom_boxplot() expects 1 discrete and 1 continuous

Unmapped parameters

geom_point() accepts additional arguments such as the colour, the transparency (alpha) or the size
possible to define them to a fixed value without mapping them to a variable.

ggplot(iris) +
  geom_point(aes(x = Petal.Width, y = Petal.Length), # end of aes()
             colour = "blue", alpha = 0.6, size = 3)

important

Note that parameters defined outside the aesthetics aes() are applied to all data.

Mapping aesthetics

colour

colour, alpha or size can also be mapped to a column in the data frame.
We can for example attribute a different colour to each species.

ggplot(iris) +
  geom_point(aes(x = Petal.Width, y = Petal.Length,
                 colour = Species), alpha = 0.6, size = 3)

important

Note that the colour argument is now inside aes() and must refer to a column in the dataframe.

Mapping aesthetics

shape

Mapping `shape` and `colour` to `Species`

ggplot(iris) +
  geom_point(aes(x = Petal.Width, y = Petal.Length, shape = Species, colour = Species),
             alpha = 0.6, size = 3)

`ggplot()` is passing aesthetics to all geoms

colour in geom inherited

mtcars %>% 
  ggplot(aes(x = wt, y = mpg, 
             colour = factor(am))) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE)

colour only for `geom_point()`

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point(aes(colour = factor(am))) +
  geom_smooth(method = "lm", se = FALSE)

Labels

It is easy to adjust axis labels and the title using the `labs()` function

ggplot(iris) +
  geom_point(aes(x = Petal.Width, y = Petal.Length, colour = Species),
             alpha = 0.6, size = 3) +
  labs(x = "Width", y = "Length",
       colour = "flower",
       title = "Iris dataset", subtitle = "petal measures",
       tag = "A", caption = "Fisher, R. A. (1936)")

Histograms

ggplot(iris) +
  geom_histogram(aes(x = Petal.Length, fill = Species),
                 alpha = 0.8, bins = 30)

Tip

The default bin value is 30 and will be printed out as a warning.
Specify your own to avoid the warning.

Density plot

The density is the count divided by the total number of occurences.

ggplot(iris) +
  geom_density(aes(x = Petal.Length, fill = Species),
               alpha = 0.6)

Overlaying plots

Density plot and histogram

ggplot(iris) +
  geom_histogram(aes(x = Petal.Length, y = stat(density)), fill = "darkgrey", binwidth = 0.1) +
  geom_density(aes(x = Petal.Length, fill = Species, colour = Species), alpha = 0.4) +
  theme_classic()

Stat functions

transform data

variables call with stat(var) are intermediate values calculated by ggplot2 using stat functions
geom uses a stat function to transform the data:
- geom_histogram() uses stat_bin()
- for frequency: (y = stat(count / max(count)))
- stat variable used in density plots: stat(density).
- stat_identity is used in geom_col() (no transformation)

Barcharts

categorical variables

`geom_bar()`

geom_bar() counts the number of values in each category
geom_bar() uses stat_count() (creates count column)

ggplot(iris) +
  geom_bar(aes(x = Species))
# or: geom_bar(aes(x = Species, y = stat(count)))

Barcharts

categorical variables

`geom_col()`

geom_col() uses stat_identity() leaving the data as is.
The y aesthetic is mandatory for geom_col()
Using geom_bar() with stat = "identity" will let geom_bar() to behave like geom_col()

ggplot(iris) +
  geom_col(aes(x = Species,
               y = Petal.Length))
#ggplot(iris) +
#  geom_bar(aes(x = Species, y = Petal.Length),
#           stat = "identity")

Stacked barchart

categorical variables

Let’s use the mtcars dataset now.

force factor to convert continuous -> discrete

mtcars %>%
  ggplot() +
  geom_bar(aes(x = factor(cyl),
               fill = factor(gear)))

Dodged barchart (side by side)

categorical variables

Do not stack the barcharts but adjust the horizontal position.

mtcars %>%
  mutate(cyl = factor(cyl), gear = factor(gear)) %>%
  ggplot() +  # position_dodge2 from v3.0 preserves single or total
  geom_bar(aes(x = cyl, fill = gear), position = position_dodge2(preserve = "single"))

Stacked barchart for proportions

categorical variables

Let’s stack the barcharts but show proportions.

mtcars %>%
  mutate(cyl = factor(cyl), gear = factor(gear)) %>%
  ggplot() +
  geom_bar(aes(x = cyl, fill = gear), position = "fill")

Stacked barchart for proportions

pie charts

We can easily switch to polar coordinates:

mtcars %>%
  mutate(cyl = factor(cyl), gear = factor(gear)) %>%
  ggplot() +
  geom_bar(aes(x = cyl, fill = gear), position = "fill") +
  coord_polar()

Boxplot

IQR, median

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

Boxplot

dodges by default

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

Customising the colours

manual definition

using scale_fill_manual() and scale_color_manual()
not handy as you must provide as much colours as they are groups

ggplot(mtcars) +
  geom_boxplot(aes(x = factor(cyl), y = mpg, fill = factor(am), color = factor(am))) +
  scale_fill_manual(values = c("red", "lightblue")) +
  scale_color_manual(values = c("purple", "blue"))

Predefined colour palettes

library(RColorBrewer)
par(mar = c(0, 4, 0, 0))
display.brewer.all()

Custom colours

using brewer

ggplot(mtcars) +
  geom_boxplot(aes(x = factor(cyl), y = mpg,
                   fill = factor(am), colour = factor(am))) +
  scale_fill_brewer(palette = "Pastel2") +
  scale_colour_brewer(palette = "Set1")

Colour gradient

The default gradient generated by ggplot2 is not very good…

ggplot(mtcars, aes(x = wt, y = mpg, colour = hp)) +
  geom_point(size = 3)

Colour gradient

Use the viridis palette instead.

ggplot(mtcars, aes(x = wt, y = mpg, colour = hp)) +
  geom_point(size = 3) +
  scale_colour_viridis_c()

Colour gradient

(and discrete variables)

viridis

5 different scales
viridis is colour blind friendly and nice in b&w
in ggplot2 since v3.0

cylinders example

ggplot(mtcars, 
       aes(x = wt, y = mpg, 
           colour = factor(cyl))) +
  geom_point(size = 3) +
  scale_colour_viridis_d()

building a ggplot step by step

by Gina Reynolds

Trump / Obama employee salaries {.box-8 offset-2 .icon .intro .outline}

Facets

facet_wrap()

Facets

facet_wrap()

Facets

free scales

Facets

`facet_grid()` to lay out panels in a grid

Facets

`facet_grid()` cont.

Exporting

interactive or passive mode

right panel

Using the Export button in the Plots panel

Rmarkdown reports

If needed, adjust the chunk options:
- size: fig.height, fig.width
- ratio: fig.asp…
- others

ggsave

save the ggplot object, 2nd argument
guesses the type of graphics by the extension

ggsave("my_name.png", p, width = 60, height = 30, units = "mm")
ggsave("my_name.pdf", p, width = 50, height = 50, units = "mm")

Extensions

ggplot2 introduced the possibility for the community to contribute and create extensions.

They are referenced on a dedicated site

plot your data!

Anscombe ** 10

never trust summary statistics alone; always visualize your data Alberto Cairo

source: Justin Matejka, George Fitzmaurice Same Stats, Different Graphs…

Missing features

geoms list here

geom_tile() heatmap
geom_bind2d() 2D binning
geom_abline() slope

stats list here

stat_ellipse()
stat_summary() easy mean 95CI etc.
geom_smooth() linear/splines/non linear

plot on multi-pages

ggforce::facet_grid_paginate() facets
gridExtra::marrangeGrob() plots

positions list here

position_jitter() random shift
quasirandom() is better

coordinate / transform

coord_cartesian() for zooming in
coord_flip() exchanges x & y
scale_x_log10() and y
scale_x_sqrt() and y

customise theme elements

legend & guide tweaks
major/minor grids
font, faces
margins
labels & ticks
strip positions
see live examples of pre-built themes

Programming

`Sepal.Length` is not exposed

iris_plot <- function(flower) {
  ggplot(iris, aes(x = Species, y = flower)) +
  geom_violin() +
  ggbeeswarm::geom_quasirandom()
}
iris_plot(flower = Sepal.Length)

Error in FUN(X[[i]], ...): object 'Sepal.Length' not found

strings does not help

iris_plot <- function(flower) {
  ggplot(iris, aes(x = Species, y = flower)) +
  geom_violin() +
  ggbeeswarm::geom_quasirandom(groupOnX = TRUE)
}
iris_plot(flower = "Sepal.Length")

Solution: tidyeval

quosures evaluated by `curly-curly`

iris_plot <- function(flower) {

  ggplot(iris, aes(x = Species, y = {{flower}})) +
  geom_violin() +
  ggbeeswarm::geom_quasirandom(groupOnX = TRUE)
}
iris_plot(flower = Sepal.Length)

evaluate in context, `{{}}` is a shortcut for

enquo() create a quosure, name with its env of origin
!! is bang bang that evaluate name in the appropriate context

build your plot step by step

using `esquisse`

RStudio add-in
by dreamRs

example

Before we stop

Acknowledgments

Thinkr (Colin Fay)
DreamRs (Victor Perrier, Fanny Meyer)
Hadley Wickham

Art

by Marcus Volz

A compilation of some of my gifs created with #rstats #ggplot2 #gganimate #tweenr https://t.co/nCppSOZv4W
— Marcus Volz (@mgvolz) 4 avril 2017

ggplot2

Learning objectives

You will learn to:

Introduction

ggplot2

Idea: split a graph into layers

Simple example

Wickham 2007

dataset

aesthetics

geometric object

Faceting

aka trellis or lattice plots

Split by the shape

Redundancy

Layers

Data

Motivation for this layered system

Geometric objects

geoms define the type of plot which will be drawn

geom_point()

geom_line()

geom_bar()

geom_boxplot()

geom_histogram()

geom_density()

Cheatsheet

The dataset

Convert iris as a tibble

Tip

Your first plot

Draw your first plot

Layers and pipes

Warning

ggplot1

ggplot1: original syntax

ggplot1 with the pipe

ggplot2

Mapping aesthetics

definitions

aesthetics requirements

Unmapped parameters

important

Mapping aesthetics

colour

important

Mapping aesthetics

shape

Mapping shape and colour to Species

ggplot() is passing aesthetics to all geoms

colour in geom inherited

colour only for geom_point()

Labels

It is easy to adjust axis labels and the title using the labs() function

Histograms

Tip

Density plot

Overlaying plots

Density plot and histogram

Stat functions

transform data

Barcharts

categorical variables

geom_bar()

Barcharts

categorical variables

geom_col()

Stacked barchart

categorical variables

force factor to convert continuous -> discrete

Dodged barchart (side by side)

categorical variables

Stacked barchart for proportions

categorical variables

Stacked barchart for proportions

pie charts

Boxplot

IQR, median

Boxplot

dodges by default

`ggplot2`

`geom_point()`

`geom_line()`

`geom_bar()`

`geom_boxplot()`

`geom_histogram()`

`geom_density()`

Convert iris as a `tibble`

`ggplot1`

`ggplot1`: original syntax

`ggplot1` with the pipe

`ggplot2`

Mapping `shape` and `colour` to `Species`

`ggplot()` is passing aesthetics to all geoms

colour only for `geom_point()`

It is easy to adjust axis labels and the title using the `labs()` function

`geom_bar()`

`geom_col()`

`facet_grid()` to lay out panels in a grid

`facet_grid()` cont.

`Sepal.Length` is not exposed

quosures evaluated by `curly-curly`

evaluate in context, `{{}}` is a shortcut for

using `esquisse`