How to add median values on top of a grouped density plot?-CodePudding

I apologise if this has been asked before. I am trying to add median values to the peak of a grouped density plot (example below).

library(dplyr)
library(forcats)

Catalan_elections %>%
  mutate(YearFct = fct_rev(as.factor(Year))) %>%
  ggplot(aes(y = YearFct))  
  geom_density_ridges(
    aes(x = Percent, fill = paste(YearFct, Option)), 
    alpha = .8, color = "white", from = 0, to = 100
  )  
  labs(
    x = "Vote (%)",
    y = "Election Year",
    title = "Indy vs Unionist vote in Catalan elections",
    subtitle = "Analysis unit: municipalities (n = 949)",
    caption = "Marc Belzunces (@marcbeldata) | Source: Idescat"
  )  
  scale_y_discrete(expand = c(0, 0))  
  scale_x_continuous(expand = c(0, 0))  
  scale_fill_cyclical(
    breaks = c("1980 Indy", "1980 Unionist"),
    labels = c(`1980 Indy` = "Indy", `1980 Unionist` = "Unionist"),
    values = c("#ff0000", "#0000ff", "#ff8080", "#8080ff"),
    name = "Option", guide = "legend"
  )  
  coord_cartesian(clip = "off")  
  theme_ridges(grid = FALSE)

CodePudding user response：

Edit:

Thanks for updating your question; I misunderstood and thought you wanted to highlight the medians (straightforward) but it sounds like you actually want the peaks (more complicated). I also thought that this was your code, not an example from https://cran.r-project.org/web/packages/ggridges/vignettes/gallery.html, so I didn't realise the Catalan_elections dataset was publicly available (e.g. from the ggjoy package).

Here is a more relevant solution:

library(tidyverse)
library(palmerpenguins)
library(ggridges)
#install.packages("ggjoy")
library(ggjoy)

Catalan_elections_with_max_density <- Catalan_elections %>%
  group_by(Year, Option) %>%
  na.omit() %>%
  mutate(max_density = max(density(Percent, na.rm = TRUE)$y),
         which_max_density = which.max(density(Percent, na.rm = TRUE)$y)) %>%
  mutate(which_max_x_intercept = density(Percent, na.rm = TRUE)$x[which_max_density])

Catalan_elections_with_max_density %>%
  mutate(YearFct = fct_rev(as.factor(Year))) %>%
  ggplot(aes(y = YearFct))  
  geom_density_ridges(
    aes(x = Percent, fill = paste(YearFct, Option)), 
    alpha = .8, color = "white", from = 0, to = 100,
  )  
  geom_segment(aes(x = which_max_x_intercept,
                   xend = which_max_x_intercept,
                   y = as.numeric(YearFct),
                   yend = as.numeric(YearFct)   max_density * 48),
               color = "white", size = 0.75, alpha = 0.1)  
  labs(
    x = "Vote (%)",
    y = "Election Year",
    title = "Indy vs Unionist vote in Catalan elections",
    subtitle = "Analysis unit: municipalities (n = 949)",
    caption = "Marc Belzunces (@marcbeldata) | Source: Idescat"
  )  
  scale_y_discrete(expand = c(0, 0))  
  scale_x_continuous(expand = c(0, 0))  
  scale_fill_cyclical(
    breaks = c("1980 Indy", "1980 Unionist"),
    labels = c(`1980 Indy` = "Indy", `1980 Unionist` = "Unionist"),
    values = c("#ff0000", "#0000ff", "#ff8080", "#8080ff"),
    name = "Option", guide = "legend"
  )  
  coord_cartesian(clip = "off")  
  theme_ridges(grid = FALSE)
#> Picking joint bandwidth of 3.16

^{Created on 2021-12-14 by the reprex package (v2.0.1)}

NB. I don't really understand how the scaling works in geom_density_ridges(), so I used "max_density * a constant" to get it approximately correct. Depending on your use-case you will need to adjust the constant or work out how the peak density relates to the y coordinates of the plot.

Original answer:

I don't have your dataset "Catalan_elections", so here is an example using the palmerpenguins dataset:

library(tidyverse)
library(palmerpenguins)
library(ggridges)

penguins %>%
  na.omit() %>%
  mutate(YearFct = fct_rev(as.factor(year))) %>%
  ggplot(aes(x = bill_length_mm, y = YearFct, fill = YearFct))  
  geom_density_ridges(
    alpha = .8, color = "white", from = 0, to = 100,
    quantile_lines = TRUE, quantiles = 2
  )  
  labs(
    x = "Vote (%)",
    y = "Election Year",
    title = "Indy vs Unionist vote in Catalan elections",
    subtitle = "Analysis unit: municipalities (n = 949)",
    caption = "Marc Belzunces (@marcbeldata) | Source: Idescat"
  )  
  scale_y_discrete(expand = c(0, 0))  
  scale_x_continuous(expand = c(0, 0))  
  scale_fill_cyclical(
    breaks = c("1980 Indy", "1980 Unionist"),
    labels = c(`1980 Indy` = "Indy", `1980 Unionist` = "Unionist"),
    values = c("#ff0000", "#0000ff", "#ff8080", "#8080ff"),
    name = "Option", guide = "legend"
  )  
  coord_cartesian(clip = "off")  
  theme_ridges(grid = FALSE)
#> Picking joint bandwidth of 1.92

^{Created on 2021-12-13 by the reprex package (v2.0.1)}