Graphics: Customization

Prologue

In our previous session session, we looked at creating simple graphs with no aesthetics or additional annotations whatsoever.
Sometimes we want to tweak the non-data part of our graph, or should I say “the little things”.
Nevertheless, it is always the little things that make the biggest difference. Therefore, I stongly believe in also giving great attention to small details.
Customazing one’s graph is often for the intention of presentation or publication. Oftertimes, we would plot quick and basic graphs as previously for exploration or exploratory analysis.
We will be using the ggplot2 for plotting purposes here.
We will be using the iris dataset availabe in base R here.

library(ggplot2)

data(iris)
class(iris); dim(iris); str(iris)

## [1] "data.frame"

## [1] 150   5

## 'data.frame':    150 obs. of  5 variables:
##  $ Sepal.Length: num  5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##  $ Sepal.Width : num  3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##  $ Petal.Length: num  1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##  $ Petal.Width : num  0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##  $ Species     : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...

Title and axis labels

Pass the relevant arguments into the labs() function.
The title argument specifies your main title.
The subtitle argument specifies the smaller-front title just below your main title.
The x argument specifies your x-axis label.
The y argument specifies your y-axis label.

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

# Ornamented
ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length)) +
  labs(title="Iris Dataset", subtitle="Correlation between septal length and petal length", x="Sepal length", y="Petal Length")

The main and subtitle are left-aligned by default.
Pass the hjust argument to the element_text() function to configure the alignment.
The plot.title and plot.subtitle objects refer to the main and subtitle, respectively.

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length)) +
  labs(title="Iris Dataset", subtitle="Correlation between septal length and petal length", x="Sepal length", y="Petal Length") +
  theme(plot.title=element_text(hjust = 0.5), plot.subtitle=element_text(hjust = 0.5))

Tick labels

This would be relevant for categorial variables mapped to the x-axis.
You may prefer labels other than the default ones in your data frame.
Pass the labels argument to the scale_x_discrete() function for this purpose.

# Bare
ggplot(data=iris) +
  geom_boxplot(mapping=aes(x=Species, y=Sepal.Length))

# Ornamented
ggplot(data=iris) +
  geom_boxplot(mapping=aes(x=Species, y=Sepal.Length)) +
  scale_x_discrete(labels=c("Setosa species", "Versicolor species", "Virginica species"))

Tick values and intervals

Or more appropriately called axis scale and intervals.
This would be relevant for continuous variables mapped to either the x- or y-axis.
Pass the limits argument to the scale_x_continuous() and scale_y_continuous() functions for this purpose.
The first variable specifies the lower limit. The second variable specifies the upper limit.

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length)) +
  scale_x_continuous(limits=c(4, 9)) +
  scale_y_continuous(limits=c(1, 8))

You may want to also specify the intervals in addition to specifying the lower and upper limits of the scale.
Pass the breaks argument to the scale_x_continuous() and scale_y_continuous() functions for this purpose.
The first variable specifies the lower limit. The second variable specifies the upper limit. The third variable specifies the interval.

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length)) +
  scale_x_continuous(breaks=seq(4, 9, by=0.5)) +
  scale_y_continuous(breaks=seq(1, 8, by=0.5))

Front size

Pass the size argument to the element_text() function to configure the alignment.
The axis.text and axis.title objects refer to the tick labels and axis labels, respectively.

# Smaller front
ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length))  +
  labs(title="Iris Dataset", subtitle="Correlation between septal length and petal length", x="Sepal length", y="Petal Length") +
  theme(axis.text=element_text(size=1), axis.title=element_text(size=2), plot.subtitle=element_text(hjust=0.5, size=3), plot.title=element_text(hjust=0.5, size=4))

# Larger front
ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length))  +
  labs(title="Iris Dataset", subtitle="Correlation between septal length and petal length", x="Sepal length", y="Petal Length") +
  theme(axis.text=element_text(size=11), axis.title=element_text(size=12), plot.subtitle=element_text(hjust=0.5, size=13), plot.title=element_text(hjust=0.5, size=14))

Aside from size, there are other parameters for which you may configure. The complete list as follows.

#element_text(family = NULL, face = NULL, colour = NULL, size = NULL, hjust = NULL, vjust = NULL, angle = NULL, lineheight = NULL, color = NULL)

Color fills

We will been looking into using just 1 color here.
We will look into using >1 colours for later session when we look into adding new variables/dimensions/levels.
Pass the color argument to the geom_*() function for dots.
Pass the fill argument geom_*() function for boxes.

ggplot(data=iris) +
  geom_boxplot(mapping=aes(x=Species, y=Sepal.Length), fill="blue")

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length), color="blue")

Shapes

The default shape for dot points, such as those in the scatterplots, are round opaques ones.
The full list of shapes can be found here.
Pass the shape argument geom_*() function to specify your preferred shape.

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length), shape=23)

# Combine color with shape
ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length), shape=23, fill="blue")

- Note that, the fill argument can be used only for the point shapes 21 to 25.

Borders

You may want your bars and your points in your boxplots or scatterplots, respectively to have colored borders.
For boxplots (and also bar charts), you would need to pass the fill and color arguments into the geom_*() function where fill and color specify the fill and border, respectively.
Similarly for scatterplots, you would need to pass the fill and color arguments into the geom_*() function where fill and color specify the fill and border, respectively.

ggplot(data=iris) +
  geom_boxplot(mapping=aes(x=Species, y=Sepal.Length), fill="blue", color="red")

ggplot(data=iris) +
  geom_point(mapping=aes(x=Sepal.Length, y=Petal.Length), shape=23, fill="blue", color="red")

Background

The default background color is grey with white grids.
Use the theme_*() functions to specify which background you prefer.
There are numerous theme_*() functions. Here, we will look into just a few..
The theme_bw() yields light grids against an empty background
The theme_classic() darkens the x- and y-axes but removes the grids.
The theme_linedraw() hightlights the edge of the graph and retains the grids.

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

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

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

Flipped graphs

Useful for some bar charts and boxplots especially when the variables are ordered in some way, i.e. in ascending or descending order.
Switches the position of x- and y-axes.
The coord_flip() is useful for this purpose.
Example of a flipped bar chart as follows.

# Calculate mean of sepal length for each species
collapsed <- aggregate(Sepal.Length ~ Species, data=iris, mean)

# Not flipped
ggplot(data=iris) +
  geom_bar(mapping=aes(x=Species, y=Sepal.Length), stat="identity")

# Flipped
ggplot(data=iris) +
  geom_bar(mapping=aes(x=Species, y=Sepal.Length), stat="identity") +
  coord_flip()

Notice that the tick labels are automatically rotated to a horizonal position.
Example of a flipped boxplot as follows.

ggplot(data=iris) +
  geom_boxplot(mapping=aes(x=Species, y=Sepal.Length)) +
  coord_flip()

Summary

Remember that you may combine customization for different features into a single function, e.g. you may customize the front, alignment, color etc. all in one function call.
These are just some of the features for customization that I can think right off the top of my head.
As you plot your own graphs, you may realize additional features that you may wish to customize.