Search code examples
rggplot2

dumbell plot with more than one start and end point

I have the following dumbbell plot which has more than one start and end point for certain groups:

library(tidyverse)
df <- structure(list(company = c("A", "A", "A", "A", "B", "B", "C", 
                           "C", "D", "D", "E", "E", "F", "F", "G", "G", "H", "H", "I", "I", 
                           "J", "J", "J", "J"), variable = c("customers", "potential_customers", 
                                                             "customers", "potential_customers", "customers", "potential_customers", 
                                                             "customers", "potential_customers", "customers", "potential_customers", 
                                                             "customers", "potential_customers", "customers", "potential_customers", 
                                                             "customers", "potential_customers", "customers", "potential_customers", 
                                                             "customers", "potential_customers", "customers", "potential_customers", 
                                                             "customers", "potential_customers"), date = structure(c(19840, 
                                                                                                                     19957, 20023, 20223, 19811, 20149, 19773, 20202, 19806, 20135, 
                                                                                                                     19859, 20201, 19815, 19951, 19786, 20085, 19854, 19909, 19831, 
                                                                                                                     19977, 19823, 20089, 20123, 20223), class = "Date")), row.names = c(NA, 
                                                                                                                                                                                         -24L), class = "data.frame")
df
ggplot(df, aes(x = date, y = company)) +
  geom_line() +
  geom_point(aes(color = variable), size = 3) +
  theme(legend.position = "bottom")

enter image description here

What is the simplest approach to add the gaps for A and J in an automated way in this instance? thanks

Solution

  • Add a column to your data which groups the pairs of observations on customers and potential_customers. This column could then be mapped on the group aes, something like:

    library(tidyverse)
df |>
  dplyr::mutate(
    group = row_number(),
    .by = c(company, variable)
  ) |>
  ggplot(aes(
    x = date, y = company,
    group = interaction(company, group)
  )) +
  geom_line() +
  geom_point(aes(color = variable), size = 3) +
  theme(legend.position = "bottom")