Transition Plot in R, when we have quantitative data for change in time, visualization is straight forward but in the case of a categorical variable, it’s not as easy.

In this article, we are going to describe transition plots for categorical variables.

Funnel Chart in R-Interactive Funnel Plot »

Approach 1:-

You can use plotmat function from the diagram package.

If you are not installed, let’s install the package.

install.packages("diagram")
library(diagram)
plotmat(transition_matrix[1:3,1:3])

Bubble Chart in R-ggplot & Plotly » (Code & Tutorial) »

Approach 2:-

Let’s load the package transition plot function from Gmisc package.

library(Gmisc)
library(grid)

Let’s create a matrix for visualization,

no_boxes <- 3
transition_matrix <- matrix(NA, nrow = no_boxes, ncol = no_boxes)
transition_matrix[1, ] <- 200 * c(.5, .25, .25)
transition_matrix[2, ] <- 540 * c(.75, .10, .15)
transition_matrix[3, ] <- 340 * c(0, .2, .80)
transition_matrix

     [,1] [,2] [,3]
[1,]  100   50   50
[2,]  405   54   81
[3,]    0   68  272

Let’s load the transition plot function and fill the box names

Visualization Graphs-ggside with ggplot »

Transition plot in R

transitionPlot(transition_matrix,
               box_txt = c("First", "Second", "Third"),
               type_of_arrow = "simple",
               min_lwd = unit(1, "mm"),
               max_lwd = unit(6, "mm"),
               overlap_add_width = unit(1, "mm"))

Based on a transition plot function, visualizing time change is quick and provide more intuitive understanding. The lines indicates the transition from one particular group/level into the next.

library(RColorBrewer)
output_perc <-
  function(txt, n) sprintf("%s\n[%.0f%%]", txt, n)
box_txt <-
  cbind(mapply(output_perc,
               txt = c("First", "Second", "Third"),
               n = prop.table(rowSums(transition_matrix))*100),
        mapply(output_perc,
               txt = c("First", "Second", "Third"),
               n = prop.table(colSums(transition_matrix))*100))

transitionPlot(transition_matrix,                
box_label = c("Before", "After"), 
               box_txt = box_txt,  
              cex = 1.2,      
          type_of_arrow = "simple") 

You can add title while using main, box labels based on box_label, and customizing box text can using box_txt. The box_txt vector assumes the same text or label for both left and right boxes.

Conclusion

The function transition plot is from Gmisc-package and it’s very handy when we need categorical time change visualization. The original idea was for transitionPlot is to show the before and after impact.

Subscribe to the Newsletter and COMMENT below!

[newsletter_form type=”minimal”]

Animated Graph GIF with gganimate & ggplot »