Creating non standard plot in R

Question

I'm not very experienced with R visualizations. Maybe somebody can give some hints on creating a plot similar to the image below with R (what functions, and package to use). In the image is a pseudo-example, using actual data would be more such "bars" in the plot. Here we have a column called "Ratio" with values A B C D and another column called "Variants" with values x, y, z, q, etc.

Answer 1

Here is a way. If the data resembles the fake data below, use geom_bar and geom_text to display the percent bars with the counts in them. Then it's just a matter of embellishing the plot as in the posted picture.

color_clrs <- c(
  A = "white",
  B = "black",
  C = "black",
  D = "white"
)
fill_clrs <- c(
  A = "#1f3560",
  B = "#bfbfbf",
  C = "#f2f3f3",
  D = "#ff0000"
)

library(ggplot2)

ggplot(df1, aes(Variants, fill = Ratio)) +
  geom_bar(position = "fill") +
  geom_text(stat = "count", aes(label = after_stat(count), color = Ratio), 
            position = position_fill(vjust = 0.5), show.legend = FALSE) +
  scale_x_discrete(limit = rev) +
  scale_y_continuous(trans = "reverse") +
  scale_fill_manual(values = fill_clrs) +
  scale_color_manual(values = color_clrs) +
  coord_flip() +
  theme_classic() +
  theme(legend.position = "top",
        axis.ticks.x = element_blank(),
        axis.text.x = element_blank())

^{Created on 2023-06-13 with reprex v2.0.2}

---

Test data

set.seed(2023)
n <- 1000
df1 <- data.frame(
  Ratio = sample(LETTERS[1:4], n, TRUE),
  Variants = sample(letters[17:26], n, TRUE)
)
head(df1)
#>   Ratio Variants
#> 1     A        z
#> 2     D        s
#> 3     C        y
#> 4     A        s
#> 5     D        v
#> 6     C        r

^{Created on 2023-06-13 with reprex v2.0.2}