I recently found THIS site on the internet which I thought would be another good case-study.

As a standalone figure, it can be rendered as follows with ggtern:

#Load the data
df <- read.table(file = "~/Desktop/PPS-data.txt", header = T)

#Create the plot and store
plot <- ggtern(data = df, aes(x = Xyp, y = XO, z = XY)) + 
               geom_point(aes(fill = Root),
                          size = 4, 
                          shape = 21, 
                          color = "black") + 
               ggtitle("PPS 3-State Model") + 
               labs(fill = "Root States") + 
               theme_tern_rgbw() + 
               theme(legend.position      = c(0,1), 
                     legend.justification = c(0, 1))

#Render
plot

However, in order to distinguish the different categories, perhaps a facetted approach would be better:

#Render w/ Facetting
plot + facet_wrap(~Root)

The post PPS 3-State Model appeared first on ggtern: ternary diagrams in R.