The box-whisker plot (or a boxplot) is a quick and easy way to visualize complex data where you have multiple samples.

A box plot is a good way to get an overall picture of the data set in a compact manner.

The boxplot() function

You can use the boxplot() function to create box-whisker plots.

It has many options and arguments to control many things, such as the making it horizontal, adding labels, titles and colors.

Syntax

The syntax for the boxplot() function is:

boxplot(x,names,xlab,ylab,border,col,notch,horizontal,add,…)

Parameters

Create a Box-Whisker Plot

To get started with plot, you need a set of data to work with. Let’s consider the built-in ToothGrowth data set as an example data set.

Here are the first six observations of the data set.

# First six observations of the 'ToothGrowth' data set
head(ToothGrowth)
   len supp dose
1  4.2   VC  0.5
2 11.5   VC  0.5
3  7.3   VC  0.5
4  5.8   VC  0.5
5  6.4   VC  0.5
6 10.0   VC  0.5

To create a box plot just specify any variable of the data set in boxplot() function.

boxplot(ToothGrowth$len)

Horizontal Box Plot

You can also draw the box-plot horizontally by setting the horizontal argument to TRUE.

boxplot(ToothGrowth$len,
        horizontal = TRUE)

Notched Box Plot

The notched box plot allows you to assess whether the medians are different. If the notches do not overlap, there is strong evidence (95% confidence) their medians differ.

You add notches to a box plot by setting the notch argument to TRUE.

# Add notches to a box plot
boxplot(ToothGrowth$len,
        notch = TRUE)

Side-by-Side Box Plots

Often your data set contains a numeric variable (quantitative variable) and a factor (categorical variable). It can be quite tedious to find whether the numeric variable changes according to the level of the factor.

Information of that nature can be gained by plotting box plots side by side.

In R, you can do this by using the boxplot() function with a formula:

boxplot(x ~ f)

Here, x is the numeric variable and f is the factor.

# Creating one box plot for each factor level (dose)
boxplot(len ~ dose, data = ToothGrowth)

Grouped Box Plot

A grouped box plot is used when you have a numerical variable, several groups and subgroups.

You can create a grouped box plot by putting interaction of two categorical variables on x-axis and a numeric variable on y-axis.

# Box plot of length based on interaction of two variables (supplement and dose)
boxplot(len ~ supp*dose, data = ToothGrowth)

Change Group Names

To change names for group of boxes, use names argument.

boxplot(len ~ dose, data = ToothGrowth,
        names=c("0.5 mg","1 mg","2 mg"))

Change Colors

Use col argument to change the fill colors used for the boxes.

boxplot(len ~ dose, data = ToothGrowth,
        col = "dodgerblue1")

You can change the colors of individual boxes by passing a vector of colors to the col argument.

boxplot(len ~ dose, data = ToothGrowth,
        col = c("orange1", "dodgerblue1", "olivedrab2"))

By using the border argument, you can even change the color used for the border of the boxes.

boxplot(len ~ dose, data = ToothGrowth,
        col="lightblue1",
        border="dodgerblue3")

Adding Titles and Axis Labels

You can add your own title and axis labels easily by specifying following arguments.

boxplot(len ~ dose, data = ToothGrowth,
        main="Tooth Growth in Guinea Pigs",
        xlab="Vitamin C dose (mg/day)",
        ylab="Length of odontoblasts")

Add Means to a Box Plot

The horizontal line in the middle of a box plot is the median, not the mean.

The median alone will not help you understand if the data is normally distributed. So, you need to add mean markers on your box plot.

boxplot(len ~ dose, data=ToothGrowth,
        col="dodgerblue1")
meanval <- by(ToothGrowth$len, ToothGrowth$dose, mean)
points(meanval, col="white", pch=18)