Beautifying Sankey/Alluvial visualization using R

Question

I'm currently using Alluvial package in R to create a visualization.

Here is my data set:

https://app.box.com/s/6qju42u0cg1cmrnhyhyzmjtp59wnsn3q

Here is my code:

alluvial(fb_ad_3d[,2:3], freq=fb_ad_3d$freq,
         col = ifelse( fb_ad_3d$Response == "Yes", "skyblue1", 
                       "darkorchid1" ),xw = 0.2,alpha = 0.6,
                        gap.width=0.2,cex = 1.1, cex.axis = 1.5)

Here is the visualization:

There are two thing I really don't like:

1. The zigzag patterns on the edges of the flow connectors

2. Certain categories (Agriculture, Events, Electronics, Telecom) on the left side have been compressed making them ineligible.

Any way to improve this visualization and make it beautiful?

Cruikshank

Have you tried the ggalluvial package? – Hey

I've not yet tried that package. – Cruikshank

Answer 1

Tried out ggalluvial package. The result is much better.

Here is the code:

A_col <- "firebrick3"
B_col <- "darkorange"
C_col <- "deepskyblue3"
alpha <- 0.7

ggplot(fb_ad_3d,
       aes(weight = freq, axis1 = Category, axis2 = Response)) +
  geom_alluvium(aes(fill = Response, color = Response), 
                width = 1/12, alpha = alpha, knot.pos = 0.4) +
  geom_stratum(width = 1/6, color = "grey") +
  geom_label(stat = "stratum", label.strata = TRUE) +
  scale_x_continuous(breaks = 1:2, labels = c("Category", "Response"))     +
  scale_fill_manual(values  = c(A_col, B_col, C_col)) +
  scale_color_manual(values = c(A_col, B_col, C_col)) +
  ggtitle("Relevance of Facebook Custom List Advertising") +
  theme_minimal() +
  theme(
   axis.text.x = element_text(size = 12, face = "bold")
  )

Here is the visualization:

Cruikshank

Hi @user709413. This is cool. Thanks! Any way to get rid of the legend? I tried the usual solutions but they are not working, not sure why. Also, how to remove the grids that are not directly related to the plot? – Cryotherapy

Try to use the following: theme( legend.position = "none", panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.text.y = element_blank(), axis.text.x = element_text(size = 12, face = "bold") ) – Cruikshank

Thanks, @user709413. Just tried the new theme and it improves tremendously the visual. – Cryotherapy