Density Plot with Filled Area by Quantiles

Probability Density Plot

This is a short coding tutorial of creating a probability density plot with areas of the density plot is filled by quantile of interest. I have used the faithful data from the dataset library and ggplot2 for visualization.

Example

You can modify the code based on your need.:

library(ggplot2)

# Estimated probability density
estDensity <- density(faithful$eruptions)

# Compute the quantile values and create a factor variable for the quantiles
estQuant <- quantile(faithful$eruptions, probs = c(0.75, 0.9, 0.95))
df_dens <- data.frame(
  x = estDensity$x,
  y = estDensity$y,
  quantFct = factor(findInterval(estDensity$x, estQuant))
)

# Density plot using ggplot2
# 1. Use geom_line() with the estimated density for each x=eruptions
# 2. geom_ribbon() to fill the areas for the quantiles 
ggplot(
  data = df_dens
)+
  geom_line(
    aes(
      x=x,
      y=y
    )
  )+
  geom_ribbon(
    aes(
      x=x, ymin=0, ymax=y, fill = quantFct
    ),
    alpha = 0.25
  )+
  scale_x_continuous(
    breaks=estQuant
  )+
  scale_fill_manual(
    name = "",
    values = c("green","yellow","orange","red")
  )+
  labs(title = "Probability Density")+
  xlab("Eruptions")+ylab("Estimated Density")+
  theme_bw()+
  theme(
    axis.text.x = element_text(angle = 90),
    panel.grid = element_blank(),
    legend.position = "none"
  )

If you want more explanation or more work similar to this, just write your comments.