How to colour area under the curve in a hypnogram?

Question

I'm trying to make an hypnogram and add color to the wake and sleep stages, something like this:

This is an example of the data frame

df <- data.frame (
time = seq(1, 250),
state = c(rep ("AW", 20), rep ("QW",35), rep ("NR", 100), rep ("Rt", 40), rep("R", 55))

And I used this code

library(ggplot2)
library(ggthemes)

ggplot(data = df, aes(x=time, y= state))+
   geom_step(aes(group=1))+
   geom_line(aes(x=time, y= state, colour = state))+
   labs (title = "Subject 1", x = "Time (s)", y= "State")+
   theme_clean()

but I got this

And actually I want that to fill "AW" and "QW" with one color and "Rt" and "R" with another color

This is another example I made with inkscape

enter image description here

Answer 1

Thanks for including your expected outcome; here is one potential approach you could try:

library(tidyverse)

df <- data.frame (
  time = seq(1, 250),
  state = c(rep ("AW", 20), rep ("QW", 35), rep ("NR", 100), rep ("Rt", 40), rep("R", 55)))

df %>%
  mutate(state_numeric = case_when(state %in% c("AW", "QW") ~ 1,
                                   state %in% c("Rt", "R") ~ -1,
                                   state == "NR" ~ 0,
                                   TRUE ~ NA_integer_),
         state = factor(state, levels = c("AW", "QW", "NR", "Rt", "R"),
                        labels = c("W", "W", "NR", "R", "R"))) %>%
  ggplot(aes(x = time, y = state_numeric)) +
  geom_step(aes(group = 1)) +
  geom_line(aes(x = time, y = state_numeric, colour = state)) +
  labs(title = "Subject 1", x = "Time (s)", y = "State") +
  theme_minimal() +
  geom_area(aes(fill = state, colour = state),
            alpha = 0.3) +
  scale_y_continuous(breaks = c(-1, 0, 1), labels = c("R", "NR", "W"))

^{Created on 2024-07-08 with reprex v2.1.0}

---

Based on your comment (if I just want to color the line, do I only use geom_line? and if I want to identify every state ("AW", "QW", "NR", "Rt", "R")?), perhaps this is what you want?

library(tidyverse)

df <- data.frame (
  time = seq(1, 250),
  state = c(rep ("AW", 20), rep ("QW", 35), rep ("NR", 100), rep ("Rt", 40), rep("R", 55)))

df %>%
  mutate(state_numeric = case_when(state %in% c("AW", "QW") ~ 1,
                                   state %in% c("Rt", "R") ~ -1,
                                   state == "NR" ~ 0,
                                   TRUE ~ NA_integer_),
         state = factor(state, levels = c("AW", "QW", "NR", "Rt", "R"))) %>%
  ggplot(aes(x = time, y = state_numeric)) +
  geom_step(aes(group = 1)) +
  geom_line(aes(x = time, y = state_numeric, colour = state),
            linewidth = 2) +
  labs(title = "Subject 1", x = "Time (s)", y = "State") +
  theme_minimal() +
  scale_y_continuous(breaks = c(-1, 0, 1), labels = c("R", "NR", "W"))

^{Created on 2024-07-09 with reprex v2.1.0}