alluvial_example.R

### Messing around with alluvial plots


library(tidyverse)
library(here)

### -------------- Note on Fonts --------------------
## For these fonts to work you will need to have
## the full Myriad Pro font installed along with
## my {myriad} package. Myriad Pro is available
## from Adobe, and may already be installed on you
## computer. The {myriad} package is here:
## https://kjhealy.github.io/myriad/
## Install it with:
## remotes::install_github("kjhealy/myriad")

## If you don't have the fonts, you can comment
## out everything from here until the "Font setup ends"
## line below. Also change or remove the references
## to "Myriad Pro Condensed" in the main code below.
## With those removed, the plots will still work, but
## they'll be in Helvetica, and they won't look as good.

library(systemfonts)
clear_registry()



register_variant(
  name = "Myriad Pro SemiCondensed",
  family = "Myriad Pro",
  width = "semicondensed",
  weight = c("normal", "semibold"),
)

library(showtext)
showtext_opts(dpi = 300)
showtext_auto()

library(myriad)
import_myriad_semi()
import_myriad_condensed()

theme_set(theme_myriad_semi())
### -------------- Font Setup Ends --------------------


### Begin figures proper

## For the flow diagrams
# install.packages("ggalluvial")
library(ggalluvial)

## For the inside labels
# install.packages("ggfittext")
library(ggfittext)

## Built in to the package
data(majors)

head(majors)

## Convert df to tibble and make the labels a little nicer.
## Also there's a typo in the sample data for some reason.
majors_df <- as_tibble(majors)  |>
  mutate(
    curriculum = str_replace(curriculum, "Sculpure", "Sculpture"),
    curriculum_f = fct_na_value_to_level(as.factor(curriculum),
                                         "(Missing)"),
    semester_f = str_replace(semester, "CURR", ""),
    semester_f= str_pad(semester_f, width = 2, pad = "0")) # So it'll sort properly

###
### Make the plot
###
p1 <- majors_df |>
  ggplot(aes(x = semester_f,
             stratum = curriculum_f,
             alluvium = student, # the thing that "flows"
             fill = curriculum_f,
             label = curriculum_f)) +
  scale_fill_brewer(type = "qual",
                    palette = "Set2") +
  geom_flow(stat = "alluvium",
            lode.guidance = "frontback",
            color = "gray30") +
  geom_stratum() +
  theme(legend.position = "bottom") +
  labs(title = "Student Curriculums Across Several Semesters",
       x = "Semester",
       fill = "Curriculum")

p1

## Alluvial/Sankey plots like to be wide
ggsave(p1, file = "alluvial_basic.pdf", height = 5, width = 12)


###
### Experimenting a bit more
###

## Better color palette
library(prismatic)
my_colors <- color(ggokabeito::palette_okabe_ito()[c(1:3, 5:8)])
my_colors_li <- clr_lighten(my_colors)

## Mess with widths
width_factor <- 0.5

p2 <- majors_df |>
  mutate(curriculum = fct_na_value_to_level(curriculum, "(Missing)"))  |>
  ggplot(aes(x = semester_f,
             stratum = curriculum_f,
             alluvium = student, # the thing that "flows:
             fill = curriculum_f,
             label = curriculum)) + ##
  geom_flow(width = width_factor,
            stat = "alluvium",
            lode.guidance = "frontback",
            color = "gray90",
            size = rel(0.2),
            alpha = 0.5) +
  geom_stratum(width = width_factor,
               alpha = 0.8,
               color = "gray90",
               linewidth = 0.2) +
  geom_fit_text(stat = "stratum",
                width = width_factor,
                family = "Myriad Pro Condensed",
                color = "gray99",
                min.size = 1) +
  scale_fill_manual(values = my_colors) +
  scale_x_discrete(expand = c(.1, 0)) +
  guides(x = guide_axis(cap = "both")) +
  theme(legend.position = "none",
        axis.line.x = element_line(color = "gray30",
                                   linewidth = 0.75),
        axis.line.y = element_blank(),
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        axis.title.y = element_blank(),
        plot.title = element_text(hjust = 0.5)) +
  labs(title = "Student Curriculums Across Several Semesters",
       x = "Semester")

p2

ggsave(p2, file = "alluvial_polished.pdf", height = 5, width = 12)
ggsave(p2, file = "alluvial_polished.png", height = 5, width = 12)