bump plot using value rather than rank

[hueyrb]

Hi David

I'm working on mountaineering data with country, year, and Nclimbers. Because Nclimbers varies hugely by year and by country, I want to use Nclimbers rather than ranks.
I used your example, and it works well!
ggplot(df, aes(year, value, color = country)) +
geom_bump()

but with my data set I get errors
"1: In compute_group(...) : 'StatBump' needs at least two observations per group."
The problem seems to be that 5 countries (Argentina, Canada, France Russia, Switzerland) had no climbers two of the three years, and they are omitted from the plot. But I have a solution (below)

Here are the data:
country <- c( "Argentina", "Canada", "China", "China", "France" , "India", "India", "India", "Italy",
"Italy", "Japan", "Japan", "Japan" , "Nepal", "Nepal" , "Russia", "S Korea", "S Korea", "Spain", "Spain", "Switzerland", "UK" , "UK" , "UK", "USA", "USA" , "USA")

year <- c(1, 3, 1, 3, 2, 1, 2, 3, 1,2, 1, 2, 3, 2, 3, 3, 2, 3, 1, 2, 1, 1, 2, 3, 1, 2, 3)

Nclimbers <- c(20, 305, 35, 805, 318, 55, 206, 993, 67, 164, 156, 444, 431, 177, 377, 319, 350,
374, 16, 350, 33, 40, 429, 1007, 27, 859, 1668)

rank <- c(8, 9, 5, 4, 6, 3, 7, 3, 2, 9, 1, 2, 5, 8, 6, 8, 4, 7, 9, 5 , 6, 4, 3, 2, 7, 1, 1)

climbers <- tibble(country, year, Nclimbers, rank)

A solution is to fill in the missing years with Nclimbers = 0
climberfill <- climbers %>% tidyr::complete(country, year, fill = list(Nclimbers = 0))

Now when I run ggplot, it works, and all countries are represented!

ggplot(climberfill, aes(year, Nclimbers, color = country)) + geom_bump()

Perhaps add an explanation to the guidelines about this warning and how to deal with it.