Overview

ggiraphAlluvial provides interactive alluvial geoms for use with ggplot2, ggalluvial, and ggiraph.

The package is designed for situations where a regular alluvial plot is useful, but the viewer also benefits from interactive features such as:

tooltips
hover highlighting
click actions

Installation

You can install the development version from your local source or GitHub.

# install.packages("pak")
pak::pak("bips-hb/ggiraphAlluvial")

Packages used in this vignette

library(ggplot2)
library(ggalluvial)
library(ggiraph)
library(ggiraphAlluvial)
#> 
#> Attaching package: 'ggiraphAlluvial'
#> The following object is masked from 'package:ggalluvial':
#> 
#>     self_adjoin

A basic alluvial plot

We start with a small example data set from ggalluvial.

data("Titanic")

df <- as.data.frame(Titanic)
head(df)
#>   Class    Sex   Age Survived Freq
#> 1   1st   Male Child       No    0
#> 2   2nd   Male Child       No    0
#> 3   3rd   Male Child       No   35
#> 4  Crew   Male Child       No    0
#> 5   1st Female Child       No    0
#> 6   2nd Female Child       No    0

A standard alluvial plot with ggalluvial might look like this:

p_static <- ggplot(
  df, 
  aes(
    axis1 = Class,
    axis2 = Sex,
    axis3 = Age,
    y = Freq
    )
  ) +
  ggalluvial::geom_alluvium(aes(fill = Survived)) +
  ggalluvial::geom_stratum() +
  ggplot2::geom_text(stat = "stratum", aes(label = after_stat(stratum))) +
  scale_x_discrete(limits = c("Class", "Sex", "Age"), expand = c(.1, .1)) +
  labs(
    title = "Titanic data as an alluvial plot",
    y = "Count"
  )

p_static

Making the plot interactive

ggiraphAlluvial adds interactive versions of selected ggalluvial geoms.

In the example below, tooltips are attached to flows and strata. Hovering over elements in the rendered widget reveals those tooltips.

p_interactive <- ggplot(
  df,
  aes(
    axis1 = Class,
    axis2 = Sex,
    axis3 = Age,
    y = Freq
  )
) +
  geom_flow_interactive(
    aes(
      fill = Survived,
      tooltip = paste(
        "Class:", Class,
        "<br>Sex:", Sex,
        "<br>Age:", Age,
        "<br>Survived:", Survived,
        "<br>Count:", Freq
      ),
      data_id = paste(Class, Sex, Age, Survived, sep = "_")
    ),
    alpha = 0.7
  ) +
  geom_stratum_interactive(
    aes(
      tooltip = after_stat(stratum),
      data_id = after_stat(stratum)
    ),
    fill = "grey85",
    color = "grey40"
  ) +
  ggplot2::geom_text(
    stat = "stratum",
    aes(label = after_stat(stratum))
  ) +
  scale_x_discrete(limits = c("Class", "Sex", "Age"), expand = c(.1, .1)) +
  labs(
    title = "Interactive alluvial plot",
    y = "Count"
  )

girafe(ggobj = p_interactive)

Tooltips and hover effects

Interactive behaviour is controlled through aesthetics recognized by ggiraph, such as tooltip and data_id, and through options passed to girafe().

girafe(
  ggobj = p_interactive,
  options = list(
    opts_hover(css = "stroke:black;stroke-width:2px;"),
    opts_tooltip(css = "background-color:white;color:black;padding:6px;border:1px solid #ccc;")
  )
)

A richer example

Interactive alluvial plots are particularly useful when the number of flows grows and a static display becomes harder to read.

df2 <- subset(df, Freq > 0)

p2 <- ggplot(
  df2,
  aes(
    axis1 = Class,
    axis2 = Sex,
    axis3 = Age,
    y = Freq
  )
) +
  geom_flow_interactive(
    aes(
      fill = Survived,
      tooltip = paste0(
        "Class: ", Class,
        "<br>Sex: ", Sex,
        "<br>Age: ", Age,
        "<br>Survived: ", Survived,
        "<br>Freq: ", Freq
      ),
      data_id = seq_len(nrow(df2))
    )
  ) +
  geom_stratum_interactive(
    aes(
      tooltip = after_stat(stratum),
      data_id = after_stat(stratum)
    ),
    fill = "grey90",
    color = "grey50"
  ) +
  scale_x_discrete(limits = c("Class", "Sex", "Age"), expand = c(.1, .1)) +
  labs(
    title = "Interactive exploration of Titanic flows",
    fill = "Survived"
  )

girafe(
  ggobj = p2,
  options = list(
    opts_hover(css = "opacity:1;stroke:black;"),
    opts_hover_inv(css = "opacity:0.25;")
  )
)

Notes

ggiraphAlluvial is intended to work alongside ggalluvial rather than replace it. The usual ggplot2 workflow still applies: build a plot with regular layer semantics, then render it with ggiraph::girafe().

In general, interactive alluvial plots work best when:

each element has a meaningful tooltip
each interactive element has a stable data_id
labels are kept short enough to remain readable

Summary

ggiraphAlluvial makes it possible to add interactivity to alluvial plots without changing the overall ggplot2 workflow. This is especially useful for exploratory graphics, dashboards, and teaching materials.

Getting started with ggiraphAlluvial