ggalluvial with one value in axis1

Question

I am supposed to make a ggalluvial with the following dataframe :

df <- data.frame(departure = c("Paris"), arrival = c("Nantes","Caen","Nice"), value = c(0.5625, 0.312, 0.125))

When I do it, as I just have one element in axis 1, it doesn't really make the 'alluvial effect'. I obtain the following result :

df %>% 
  group_by(departure) %>% 
  ggplot(aes(y = value*100, axis1 = departure, axis2 = arrival)) +
  geom_alluvium(aes(fill = arrival), width = 1/12) +
  geom_stratum(width = 1/12, fill = "white", color = "grey") +
  geom_label(stat = "stratum", aes(label = after_stat(stratum))) +
  scale_x_discrete(limits = c("departure", "arrival"), expand = c(.05, .05)) +
  scale_fill_brewer(type = "qual", palette = "Set1") +
  theme(legend.position = "none")

I tried to cheat by adding a fictive row in order to get the expected alluvial effect.

df <- rbind(df, c("-", "Nantes", 0.5625))
df <- rbind(df, c("-", "Caen", 0.262))
df$value <- as.numeric(df$value)

Is it possible to make an aesthetic alluvial plot with one axis1 ? Thank you in advance.

Answer 1

Your data structure does not lend itself well to an alluvial plot. If this is an excercise to create an alluvial plot from the data, then you have succeeded; this is what an alluvial plot with the given data should look like. If you want something that is more in keeping with the look of an alluvial plot, you could try making a graph with more separation between the left-hand labels. This requires a bit of calculation to work out the positions

library(tidygraph)
library(ggraph)

df %>%
  mutate(departure = paste(departure, arrival)) %>%
  as_tbl_graph() %>%
  mutate(label = ifelse(grepl("Paris", name), "Paris", name),
         xpos = ifelse(label == "Paris", 1, 2),
         ypos = ifelse(label == "Paris", cumsum(df$value) - df$value/2, 
                       cumsum(df$value + 0.5) - 1)) %>%
  ggraph(layout = "manual", x = ypos, y = xpos) +
  geom_edge_diagonal(aes(width = value, color = factor(to)), alpha = 0.3) +
  geom_node_tile(aes(width = rep(df$value, 2)), height = 0.1, fill = "white") +
  geom_node_text(aes(label = label), angle = 90) +
  annotate("rect", xmin = 0, xmax = 1, ymin = 0.95, ymax = 1.05, 
           fill = "white", color = "black", linewidth = 0.2) +
  annotate("text", x = 0.5, y = 1, angle = 90, label = "Paris") +
  scale_edge_width_continuous(range = c(0, 118), limits = c(0, 1)) +
  theme(legend.position = "none") +
  coord_flip()