how to build a barplot using ggplot2 with 2 sets of bars: 1st set for all observations per group; 2nd set for number of observations meeting criteria?

I have a table that looks like this

tibble (person = c ("Paul", "Paul", "Paul", "Sarah", "Sarah", "Sarah", "Alex", "Alex", "Alex"),
    salary = c(8, 10, 3, 6, 6, 6, 1, 1, 3))

how would I build a barplot with two sets of bars side-by-side that shows:

1. the total number of salaries observations per person would give bars c(3, 3, 3)
2. the number of observations with salary higher than 5 would give bars c(2, 3, 0)

I wouldn't advise trying to do this in ggplot directly. Rather, you should precompute your summaries and plot those:

library(tidyverse)

df <- tibble(person = c ("Paul", "Paul", "Paul", "Sarah", "Sarah", "Sarah", "Alex", "Alex", "Alex"),
              salary = c(8, 10, 3, 6, 6, 6, 1, 1, 3))

salary_counts <- df |> 
  summarize(
    total_salary = n(),
    high_salary = sum(salary > 5),
    .by = person
  ) |> 
  pivot_longer(cols = c(total_salary, high_salary), names_to = 'count_type', values_to = 'count')

salary_plot <- salary_counts |> 
  ggplot(data = _, aes(x = person, y = count, fill = count_type)) +
  geom_col(position = 'dodge')
print(salary_plot)