Adding multiple legend to combined graph ggplot2

Question

I have such type of the data:

head(data,10)
   Ratio Wealthness Cash Quality Variants
1      A          A    E       G        r
2      D          B    A       C        u
3      C          B    C       E        s
4      A          C    E       A        s
5      D          B    A       D        y
6      C          B    D       D        x
7      B          C    D       D        v
8      D          C    C       D        y
9      B          C    B       A        x
10     A          A    D       G        u

And created a combined plot adapting the solution by Rui Barradas provided to my previous question :

library(ggplot2)
library(tidyr)

color_clrs <- c(
    A = "white",
    B = "black",
    C = "black",
    D = "white",
    E = "black",
    F = "white",
    G = "white"
)

fill_clrs <- c(
    A = "#1f3560",
    B = "#B0C4DE",
    C = "#f2f3f3",
    D = "#ff0000",
    E = "#A9A9A9",
    F = "#B22222",
    G = "#1E90FF"
)

ggplot(data %>% pivot_longer(-Variants), aes(Variants, fill = value)) +
    geom_bar(position = "fill") +
    geom_text(stat = "count", aes(label = after_stat(count), color = value), 
              position = position_fill(vjust = 0.5), show.legend = FALSE) +
    facet_wrap(~name) +
    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())

The output:

My question how to make separate legend with colors and groups (A-E) to each column? It is not real data example in this question, in reality some groups can be 1 to 1000 and it's even difficult to spot color in such cases. Or it can be that only to one time some letter appeared for all variants under the variable, so it will be more understandable if there would be seen what groups can be for that variable.

the code you shared, with the data from Rui Barradas's answer gives one plot, not the four in your picture. Something has gone wrong somewhere. Please explain!^_^ — Mark, Jun 19 '23 at 10:01
You can have one plot per column, with its legend, and then [combine](https://stackoverflow.com/questions/1249548/side-by-side-plots-with-ggplot2) the [plots](https://stackoverflow.com/questions/68925650/how-do-i-combine-multiple-plots-in-one-graph). — Rui Barradas, Jun 19 '23 at 10:11

M Aurélio · Answer 1 · 2023-06-19T12:40:53.073

You can transform your plot in a function (is the same code that You wrote, I just removed facet_wrap and pivot_longer parts), apply it in every subset with purrr::map2 and then combine all the plots from the list with patchwork::wrap_plots


library(patchwork)

ind_plot <- function(data, plot_titles)
  data %>% 
  ggplot(aes(Variants, fill = value)) +
  geom_bar(position = "fill") +
  geom_text(stat = "count", aes(label = after_stat(count), color = value), 
            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()) + 
  labs(title = plot_titles)

  data %>% 
  pivot_longer(-Variants) %>% 
  split(.$name) %>% 
  purrr::map2(names(.), ind_plot) %>% 
  patchwork::wrap_plots()

^{Created on 2023-06-19 with reprex v2.0.2}

Adding multiple legend to combined graph ggplot2

1 Answers1