thilo klein

Overview

1. Introduction: Plotting with ggplot2
   • A simple function: qplot
   • A robust function: ggplot
2. Scatterplot
3. Faceting
   • Facet wrap
   • Facet grid
4. Title and axis
5. Histograms
6. Boxplots and violin plots
7. Output and saving

1. Introduction: Plotting with ggplot2

First, install and load the ggplot2 package

install.packages("ggplot2")
library(ggplot2)

For this session, we will explore the iris data that is already pre-loaded in R.

head(iris)

You can read about the data by typing

?iris

1.1. A simple function: qplot

The general formula is qplot(x, y, ...). To produce a basic scatter plot, type

qplot(x=Sepal.Length, y=Petal.Length, data=iris)   

Adding color per species and size depending of petals width

qplot(Sepal.Length, Petal.Length, data=iris,
      color=Species, size=Petal.Width, 
      xlab="Sepal", ylab="Petal", main="Iris dataset")

We also add a title and labels for the x and y-axis, using main, xlaband ylab.

To start at 0 for the y-axis: add qplot(...,ylim=c(0,35))
This set lower and upper bounds for y axis.
You can do the same for the x-axis

qplot(Sepal.Length, Petal.Length, data=iris, 
      color=Species, size=Petal.Width, alpha=I(0.7), 
      xlab="Sepal Length", ylab="Petal Length", main="Iris dataset")

By setting the alpha of each point to 0.7, we reduce the effects of overplotting.

1.2. A robust function: ggplot

General formula:
ggplot(data, aes(x,y)) + geom_*()
ggplot begins a plot that you finish by adding layers to, using geom(). ggplot provides more control than qplot().

• aes: aesthetic, visual properties of the graph
  • options aes: color, fill, shape, size
• geom: graphical property
  • geom_line; geom_bar; geom_histogram
  • geom_chart; geom_hex, geom_c(point,line) etc. You can specify for each geom the aesthetic mappings, and a default stat and position adjustment: geom_*(aes(color=, fill=, size=....))
• additional elements :
  • You can add a smoothing trend : + geom_smooth (method="lm")
  • You can change the background (Themes): + theme_bw(), `+theme_classic()

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

From now on, we will only work using ggplot.

Back to top

2. Scatterplot

ggplot(diamonds, aes(x=carat, y=price)) + geom_point()

ggplot(diamonds, aes(x=carat, y=price, color=clarity)) + geom_point()

Adding color: the color of the points is determined by the clarity of the diamonds.

ggplot(diamonds, aes(x=carat, y=price, color=cut)) + geom_point()

Here we changed the color by the parameter “cut” of the dataset diamonds.

ggplot(diamonds, aes(x=carat, y=price, color=clarity, size=cut)) + geom_point()

Here we changed the size of the points according to the parameter cut.

Back to top

3. Faceting

• Divide your plot up into multiple plots, one for each level.
  
• facet_*() function

3.1. Facet wrap

Write a tilde (~), and then the attribute we would like to divide the plots by (here “clarity”)

ggplot(diamonds, aes(x=carat, y=price, color=cut)) + geom_point() + facet_wrap(~ clarity)

3.2. Facet grid

To divide your graph based on two different attributes: facet_grid()
Here we could use "color ~ clarity". The tilde (~) means “explained by.”

ggplot(diamonds, aes(x=carat, y=price, color=cut)) + geom_point() + facet_grid(color ~ clarity)

Back to top

4. Title and axis

Title: add to the end of the line of code ggtile()

ggplot(diamonds, aes(carat,price)) + geom_point() + ggtitle("My scatter plot")

Axis label: xlab and ylab

ggplot(diamonds, aes(x=carat, y=price)) + geom_point() + ggtitle("My scatter plot") + xlab("Weight (carats)")

To limit the range of the x or the y axes: xlim or ylim

ggplot(diamonds, aes(x=carat, y=price)) + geom_point() + ggtitle("My scatter plot") + xlab("Weight (carats)") + xlim(0, 2)

This code limits the diamonds weight from 0 to 2 carats.

Another possibility is to put one of the axes on a log scale. You can do this with the scale_y_log10() function:

ggplot(diamonds, aes(x=carat, y=price)) + geom_point() + ggtitle("My scatter plot") + xlab("Weight (carats)") + scale_y_log10()

Back to top

5. Histograms

Using the data set diamonds, we visualize the density of the distribution of price:

ggplot(diamonds, aes(x=price)) + geom_histogram()

You notice that we do not precise the y axis in the aesthetic. Indeed, using histograms to visualize the density of distribution, y is directly equal to the density within each bin!

You can change the width of each bin as an option, using binwidth inside the geom_histogram() function:

ggplot(diamonds, aes(x=price)) + geom_histogram(binwidth=2000)

And we can add all the options that we used with a basci scatter plot. For instance, the facet_wrap() function:

ggplot(diamonds, aes(x=price)) + geom_histogram(binwidth=200) + facet_wrap(~ clarity)

Here, all the subplots share the same y-axis: this makes it hard to interpret the frequencies…
Some subplots have far more points than others.

To have different y-axis for each subplot, add scale="free_y"

ggplot(diamonds, aes(x=price)) + geom_histogram(binwidth=200) + facet_wrap(~ clarity, scale="free_y")

The same function can be used differentiating for the x-axis.

To make a stacked histogram based here on the clarity attribute: Try adding the fill aesthetic within the aesthetic of the ggplot() function.

ggplot(diamonds, aes(x=price, fill=clarity)) + geom_histogram()

Back to top

6. Boxplots and violin plots

Boxplots are useful to compare multiple distributions. Compare the distribution of the price within each color using geom_boxplot():

ggplot(diamonds, aes(x=color, y=price)) + geom_boxplot()

Put the y-axis on a log scale: we get a better sense of how the distribution of price differs across multiple colors.

ggplot(diamonds, aes(x=color, y=price)) + geom_boxplot() + scale_y_log10()

A box plot will always look like a square. We can instead view the distribution as a density using a “violin plot”.

ggplot(diamonds, aes(x=color, y=price)) + geom_violin() + scale_y_log10()

ggplot(diamonds, aes(x=color, y=price)) + geom_violin() + scale_y_log10() + facet_wrap(~ clarity)

We used the facet_wrap() function to display multiple subplots.

Back to top

7. Output and saving

Run your line of code and save it to a variable. For instance, call it p for plot:

p = ggplot(diamonds, aes(x=carat, y=price)) + geom_point()

Save that plot to a file

ggsave(filename="diamonds.png", p)  ## as a png
ggsave(filename="diamonds.pdf", p)  ## as a pdf
ggsave(filename="diamonds.jpeg", p) ## as a jpeg

One useful shortcut is that if you just displayed a plot, like in a line like this:

ggplot(diamonds, aes(x=carat, y=price)) + geom_point()
ggsave("diamonds.png")

Then ggsave() will know to save that plot by default when you perform ggsave - you don’t even have to tell it which plot you’re saving.

Back to top