Materials

Script

  1. Click here to download the script! Save the script to the scripts folder (R Project folder).

  2. Load your script in RStudio. To do this, open RStudio and click the files window and select the scripts folder and then this script.

Let’s get started plotting!

Read data

We’ll start with the ‘trees’ dataset, which is built into R. It describes the girth, height, and volume of 31 felled black cherry trees.

  # Data ----------------------

# look at description of built in dataset
?trees      

# Show the structure of the trees dataframe
str(trees)   
## 'data.frame':    31 obs. of  3 variables:
##  $ Girth : num  8.3 8.6 8.8 10.5 10.7 10.8 11 11 11.1 11.2 ...
##  $ Height: num  70 65 63 72 81 83 66 75 80 75 ...
##  $ Volume: num  10.3 10.3 10.2 16.4 18.8 19.7 15.6 18.2 22.6 19.9 ...
 # Show the first few observations of the trees dataframe
head(trees)  
##   Girth Height Volume
## 1   8.3     70   10.3
## 2   8.6     65   10.3
## 3   8.8     63   10.2
## 4  10.5     72   16.4
## 5  10.7     81   18.8
## 6  10.8     83   19.7

Basic plots

While we will spend most of today focusing on using ggplot2 (a very powerful data visualization package in the tidyverse), you don’t always need publication quality graphs and for simple data exploration it can be quicker to know how to make plots in base R.

R’s basic “plot()” function takes an “x” argument (defining coordinates on an x axis) and a “y” argument (defining coordinates on a y axis).

Here is an example of a scatterplot in R:

  # Basic plotting ----------------------

plot(x = trees$Girth, y = trees$Volume)    # use R's built-in "trees" dataset: ?trees

# or

plot(trees$Girth, trees$Volume)

Remember the $ in base R refers to a specific column in the dataset and that we don’t have to specify ‘x =’ and ‘y=’.

Change Plot Type

Because we’re exploring different ways of plotting, it may be useful to look at different plot types.The default plot type for two quantitative variables is points (classic scatterplot), but you can change it to lines or both points and lines (or others) by using the type= option

# line graph
plot(trees$Girth, trees$Volume,
     type = "l")   # lines

Use the help shortcut in your console to look up what other types of plots you can specify in the plot argument. (Hint don’t forget to specify which package)

# ?base::plot

Viewing multiple plots at once

It can also be useful to include multiple plots in the same image.

We can do this using the par() function (graphical parameters), which has arguments that control just about every aspect of a plot in R.

  # Viewing multiple plots ----------------------

?par

par()   # view the default graphical parameters (can be kind of overwhelming!)

We could change the “mfrow” parameter from c(1,1) to c(2,2): this means that we can fit four plots into a single window.

# change plotting window

# specify graphical parameters
par(mfrow=c(2,2)) 

# create plots to fill plotting window
plot(trees$Girth, 
     trees$Volume)             # point

plot(trees$Girth, 
     trees$Volume, 
     type="l")   # lines

plot(trees$Girth, 
     trees$Volume,
     type="b")   # both

plot(trees$Girth, 
     trees$Volume, 
     type="o")   # both with connected lines

An even easier solution is to use the convenience function layout(). An example using layout() is below.

# layout

# Use "layout" to define a 2 row x 2 column matrix with elements 1, 2, 3, and 4.
# This divides the image into four sections and then fills these with the plot function
layout(matrix(1:4, nrow=2, ncol=2))

# create plots to fill plotting window
plot(trees$Girth, 
     trees$Volume)             # point

plot(trees$Girth, 
     trees$Volume, 
     type="l")   # lines

plot(trees$Girth, 
     trees$Volume, 
     type="b")   # both

plot(trees$Girth, 
     trees$Volume, 
     type="o")   # both with connected lines

Whenever you use layout() or par(), the graphics window will retain this layout for all future plots. To start over (and return to the default graphical parameters), use graphics.off() to reset the plot. For example:

plot(trees$Girth, 
     trees$Volume)             ## The plot is still in 4 parts

graphics.off()                                  ## now the plot is reset!

Change Plot Symbol

We can also change the type of points used when plotting using the pch= option. For example, we plot three different shape options below:

  # Plot symbols ----------------------

# Use layout to define a 3 row x 1 column matrix with elements 1, 2, and 3.
# This divides the image into three sections and then fills these with the plot function
layout(matrix(1:3, 
              nrow = 3, 
              ncol = 1))

# pch: 'plotting character' changes the type of point that is used (default is an open circle)!
plot(trees$Girth, 
     trees$Volume, 
     pch = 19)     # filled point

plot(trees$Girth, 
     trees$Volume, 
     pch = 2)      # open triangle

plot(trees$Girth, 
     trees$Volume, 
     pch = 11)     # star

You might want to remember a couple favorites (for example, I like to use pch = 19). Alternatively, you might consider saving a useful guide like this:

Histograms

We can also use base R to create histograms, which are useful for exploring data and checking assumptions of normality for statistical analysis.

  # Histograms ----------------------

layout(matrix(1:2, 1, 2))

## y-axis is in counts by default (total observations in each "bin")
hist(iris$Sepal.Length, 
     main = "Histogram of Sepal Length",
     xlab = "Sepal Length")

## change y-axis to proportions of the entire dataset using freq=FALSE
hist(iris$Sepal.Length, 
     freq = FALSE, 
     main = "Histogram of Sepal Length", 
     xlab = "Sepal Length")

Plot matrices

Plot

Alternatively you can also plot all columns of a dataset in a matrix using the plot() function with your dateset inside the ()

  # Plotting matrices ----------------------

plot(iris)

Pairs

The pairs() function allows for quick investigation into relationships between variables. Be careful if your data set is large (e.g., lots of columns), as this can be a slow function.

# quick invesrtigation of relationship between variables 

pairs(iris)

Correlelogram (correlations)

Another useful thing to know how to do is to create a correlelogram or a matrix depicting correlations between variables in a dataset. I used to copy and paste a super long confusing code from base R to do this but recently found that a package PerformanceAnalytics has a easy to use function that will do this for you from the link below.

Correlation plot

Install the PerformanceAnalytics package and load it to your library.

# create correlation matrix for numeric variables in the iris data
chart.Correlation(iris[,1:4], 
                  histogram = TRUE, 
                  method = "pearson")