Using R to Create Ternary Diagrams: An Example Using 2016 Presidential Polling Data

An R Tutorial by D. M. Wiig

In previous tutorials I have discussed the basics of creating a ternary plot using the ggtern package using a simple hypothetical data frame containing five values. In a subsequent tutorial I discussed the application by creating a ternary graph using election results from the British House of Commons from the last half of the 20th century. This type of plot creates a very nice visual of the effects of a third party on the election outcome.

In this tutorial I will discuss using the same technique as applied to recent polling data from the ongoing 2016 U.S. presidential campaign. Before discussing the current election campaign I am going to refresh your memory relative to using the ggtern package.

Before running the script in this tutorial make sure that the packages ggplot, ggplot2, and ggtern are loaded into your R environment. Please also note the you will need a recent version of R that is version 3.1.x or newer. A very basic graph can be easily constructed. I will the use theoretical quantities XA , XB , and XC to demonstrate a basic ternary diagram. In this simple example I will create a sample of n=5 by entering the data from the keyboard into a data frame ‘sampfile.’ To invoke the editor use the following code:

###################################################

#create a sample file of n=5

###################################################

sampfile <-data.frame(Xa=numeric(0),Xb=numeric(0),Xc=numeric(0))

sampfile <-edit(sampfile)

###################################################

This will open up a data entry sheet with three columns labeled Xa, Xb, and Xc. The number that are entered do not matter for purposes of this illustration. The table I entered is as follows:


 Xa      Xb      Xc

1 100   135   250

2 90     122    210

3 98      44     256

4 100   97     89

5 90     75     89

To produce a very basic ternary diagram with the above data set use the code segment:

##################################################

#do basic graph with sample data

##################################################

ggtern(data=sampfile, aes(x=Xa,y=Xb, z=Xc)) + geom_point()

##################################################

This produces the graph seen below:

The triangular representation of the dimensions Xa →Xb, Xc → Xa and Xb →Xc allow each case to be represented as a single point located relative to each of the three vectors. There are a large number of additions, modifications and tweaks that can be done to this basic pattern. In the next tutorial I will discuss generating a more elaborate ternary diagram using polling data from the current U.S. presidential campaign.

Thu US has a two party dominant system with several minor parties that regularly contest elections. In the current presidential election campaign there are the two major party candidates as well as two minor party candidates for the Libertarian and Green parties that are being included in the numerous public opinion polls that are being done nationally.

For purposes of this example I have added the percentages for these two minor parties together. This results in three variables that are being plotted, the percentage for Clinton (Democrat), Trump (Republican), and for the combined Johnson (Libertarian) and Stein (Green). By plotting the three variables over time on a ternary diagram we can visualize any changes in the mixture of support indicated for the candidates.

The poll data used in this project were taken from the web site RealClearPolitics.com for the time period from July 29 to August 18.¹ It should be noted that the poll numbers were not necessarily from the same polling organization for each date but all polls used were listed as being national in scope with a Clinton v. Trump v. Johnson v. Stein format.

Before working through this tutorial make sure that you have the ggplot, ggplot2, and ggtern packages loaded into your R environment.² I originally created the table shown above using Excel and then converted it into a *cvs format before importing it into R studio for analysis.³ The data can be entered directly via the R data editor as shown in the previous example. The code segment below was used to load the *csv format file:

####################################################Enter data into spreadsheet and save a a *csv file

#Load the data into a table using the read.table function

polldata <- read.table(“d:/16electiondata.csv”, header = TRUE, sep=”,”)

#Make sure the table is ok

View(polldata)

###################################################